Skip to main content

Full text of "Prehistoric Life"

See other formats


_1 60457 > 

^ DO 



Gemuendina, a specialized early Devonian chordate. The long fin-rays 
suggest the skate, a group to which it is only distantly related. Photograph 
of a reconstruction, after Broili. 






COPYRIGHT, 1959, 1947 

Second Printing 



































INDEX 313 


We have been told, perhaps too often, that "the proper study of mankind is man." 
But what of man's ancestors, man's background? Facts mean little unless they can 
be interpreted in terms of contemporaneous and previous history. Morse invented the 
telegraph, Bell the telephone, and Marconi the "wireless." Each is heralded as a won- 
der-worker, and deservedly so. But would any one of them have accomplished any- 
thing had it not been for the less spectacular labors of his predecessors in the study of 
that field of physics now commonly labeled Electricity and Magnetism? No great 
advance in knowledge has been sudden. An inspiration occurs only to one who knows 
what has previously been discovered. 

In other words, progress depends upon information accumulated in the past. 
Until recently it has been impossible for any animal, even man, to preserve records 
that might benefit future generations, but before the invention of writing, the wise 
men of prehistoric tribes handed on traditions. Even earlier than the wise men, 
animal instincts were communicated from generation to generation by mothers, with 
or without the assistance of the less responsible fathers. It is impossible to say when 
instincts, as teachable qualities, arose; but back in the distant past there were countless 
animals, orphaned before birth, whose environment furnished their only education. 
To them, life was purely an experiment. Various paths of evolutionary change were 
open to them. Their opportunities were, of course, largely controlled by their ancestry. 
Fish-roc could not produce oysters, nor would oyster-spat develop into fish. "Like 
begets like." But, although there are at present thousands of kinds of fish, and many 
kinds of oysters, it is possible to trace genealogies back to a time when there was only 
one kind of fish and one kind of oyster. Indeed, it is possible to go still further, to a 
time when there was neither fish nor oyster, only ancestral forms, more or less like 

It is the purpose of this book to trace the history of life from the time of its first 
appearance on the earth to the present. That history has been one of constant change, 
for better or for worse. Practically all Nature's changes are evolutionary in that they 
are orderly, sequential. The inorganic, as well as the organic, evolves. Erosion changes 
a plateau into mountains and then into plains. The present surface of the earth is the 
result of evolutionary processes. As will he shown, physical and organic evolution 
have gone hand in hand. It is difficult to say which has had the greater control, the 
innate qualities of living matter or the environment. Organisms are plastic, environ- 
ments rigid. Man cannot change the climates of the tropics or of the polar regions, 
but he can live in both. So far as possible, he adapts his habits to life under abnormal 
conditions, but no one will deny the fact that the environment changes him. 


Anyone familiar with the geology of the earth's surface will realize that the 
higgledy-piggledy distribution of the rocks precludes the inference that it is the result 
of design. If it had been, there would have been symmetry, such as actually exists 
beneath the crust. Geologists have so far unraveled the history of the rocks as to show 
that their present constitution, attitude, and distribution result from processes which 
have acted differently and at different times on various regions throughout a period 
of some two billion years. There have been rhythmic sequences of mountain building, 
vulcanism, emergence, and submergence, and the evolution of life seems to have been 
controlled by them. 

Under these various environmental controls, the progress of life has not been one 
grand march from Amoeba to man. Their plasticity has permittee! animals and plants 
to follow as many paths as the environment opened to them. Some led to pleasant 
places, and the groups which followed them have survived; others were ways which 
ended in the swamps of despair and extinction. Man's ancestors took the hardest trail 
of all; not, in fact, a single trail, but a series of bypaths along which they slunk for 
millions of years before they emerged upon the highroad. 

Unless one has a sense of chronology, this history will be entirely meaningless. 
If he does not know it already, the reader should at once commit to memory the 
"geological timetable" on page ix. The estimates in years are those now generally 
accepted as reasonably accurate, the basis being the rate of decay of uranium through 
the radioactive series of derivatives to the final product, lead. 

This book is the result of a constantly changing series of lectures which I have 
given at Harvard for the past seventeen years under the title of Paleontology i. When 
I came to Harvard in 1912, I knew nothing about paleontology, although I had held 
responsible positions as paleontologist previously, and was recognized as an "authority" 
in my particular field. When I found that I was expected to teach the subject, I set 
out to learn something about it. As my education progressed, I found that I could 
best interest students by telling them of the things which most interested me. If this 
book has any value, it lies in the fact that I have tried to present the general results 
of the work of hundreds of paleontologists, not specific details. 

Theoretically, education should proceed from the general to the special. That, 
however, is not the tradition in paleontology. Most paleontologists started as collectors 
of fossils. It naturally follows that taxonomy, the science of identification, differentia- 
tion, and naming of specimens, should be their first interest. Many never progressed 
beyond this stage; hence there has been a tendency to train students along narrow 
lines. It is easy to become a specialist in some particular field in a few years. This 
tendency has been fostered recently by the fact that men who know particular subjects 
are sure of getting jobs. Educational short cuts to technical positions do not appeal to 
the writer. His own experience, which, as has been explained, was an education in 
reverse, convinces him that the broader the background, the better the specialist. 




The adventitious Fossils, which are but the Exuviae of Animals have been erroneously 
thought a sort of peculiar Stones. Cotton Mather, The Christian Philosopher 

Much has been written in recent years about the early history of the earth in so 
far as it can be deduced from astronomical and physical data. The evolution of the 
world would have been futile, however, had it not been for the introduction of life. 
As to how life originated, geology unfortunately gives little information, but that the 
earth has supported life for countless millions of years is clearly shown by the remains 
of animals and plants entombed within the sedimentary rocks during their accumula- 
tion and preserved to the present time. These remains serve a twofold purpose. Not 
only do they give a clue to the history of life upon the globe, but, when properly 
studied and interpreted, they reveal much of its physical history. What prehistoric 
implements are to the archaeologist, or the inscriptions incised by ancient peoples 
upon enduring rock are to the historian, such are fossils to the geologist. Fortunately, 
their study is not nearly so difficult as that of artifacts or inscriptions, nor does it 
require so technical a training; yet it produces results of the same order of accuracy. 

To study fossils it is necessary to have some knowledge of living animals and 
plants, for fossils are either more or less perfectly preserved remains of organisms, 
or evidences of their former existence. At best they are less complete than recent 
specimens, so that, unless something is known of the modern fauna and flora, one is 
totally unable to interpret the fragments found in the rocks. Let us, then, review 
briefly some of the fundamentals of biology. 

It should be borne in mind that all matter is either organic or inorganic. If it is 
organic, in the true sense, then it is the result of life processes. Either it is, or has 
been, alive, or it was produced by something living. Chemists, unfortunately, use the 
term organic in another sense. Because carbon is one of the principal constituents of 
living organisms, they have designated as "organic chemistry" that branch of their 
subject which deals primarily with the compounds of carbon. In this book the term 
is used only in relation to what are called organisms, the chief of whose character- 
istics is "life.". What life is, no man can say, but we know that an organism has a 
power different from that of any inorganic substance in that it can take in foreign 
chemical compounds, absorb them into 'itself, and still remain structurally the same 
as before. In chemical reactions, on the other hand, if elements already in combination 
take to themselves a new element or combination of elements and unite with them, 


a product with properties decidedly different from the original is formed. Further- 
more, in addition to absorbing and excreting external matter (food) without losing 
its identity, an organism has the power of growth from within, instead of by accre- 
tion from without, and of reproduction, that is, of making new individuals like itself. 
There are two great groups of organisms, plants and animals. If one compares 
the higher animals with the herbage they eat, it seems absurd to ask what the differ- 
ence is between them; yet when one studies some of the tiny and simple living forms 
found in fresh-water, pools, it is not easy to decide to which group they belong. They 
may have the green coloring matter of a plant but the mobility of an animal. Indeed, 
there are some groups of these small beings whose classification is still uncertain, for 
they are called plants by the botanists and animals by the zoologists. Ordinarily we 
think of animals as being capable of locomotion, whereas plants are fixed in one spot; 
yet in reality many animals are fixed just as immovably as plants. We also think of 
animals as having arms and legs and a head; but many animals possess no more of 
these organs than a plant. There is, in fact, no one criterion which can be relied upon 
to distinguish the two groups. Nevertheless there are several ways in which they 
generally differ. These may be tabulated. 

PLANTS (in general) ANIMALS (in general) 

Take inorganic matter as food and con- Take organic food. 

vert it into organic. 

Contain green coloring matter (chlo- Lack chlorophyll. 


Contain cellulose. Lack cellulose. 

Lack power of locomotion. Have power of locomotion at some 

stage in life. 

Use carbon dioxide and some oxygen. Use oxygen. 

Excrete oxygen and carbon dioxide. Excrete carbon dioxide. 

Eat by absorption through the surface. Eat by a mouth and digestive canal. 

Have indefinite. growth. Have definite growth. 

Lack a nervous system. Have a nervous system. 

There are, however, exceptions to all of these general "rules." 
Life depends for existence on sunlight. Since light is a kind of kinetic or active 
energy, it would be quickly exhausted by the individual receiving it if it could not 
be transformed into potential or stored energy. Plants effect this transformation by 
the aid of chlorophyll, which enables them to make starch from inorganic compounds; 
animals obtain their energy by eating plants or other animals. It is barely possible 
that some of the earliest organisms on the globe were free-moving chlorophyll- 
possessors which both stored energy in the form of starch and made use of it in 
movement; such organisms did double duty. Since an active existence requires con- 


centrated food, there was an opportunity for division of labor and hence for the origin 
of two great groups, one which attended to the storing of energy, and a second which 
captured and used what the other had produced. Plants settle down to life on a single 
spot, manufacturing food; animals use plants or other animals as food and expend 
the energy they have acquired in motion. Plants, living on inorganic matter, have an 
assured food supply, are quickly settled in life, and merely vegetate. Animals, forced 
to hunt for their prepared food and, if necessary, to fight for it, have more points of 
contact with their environment, and are apt to change in a greater variety of ways. 

As most people know, the simplest animals are microscopic creatures, found com- 
monly in fresh water, which look like bits of transparent jelly. Some of them have 
no particular shape, or rather a constantly changing one. They possess no heads or 
bodies, arms or legs, eyes or mouths, or even any digestive tract: in fact, they are 
organless, except for a spot of slightly denser material called a nucleus. A well-known 
example of such a creature is the Amoeba, familiar to everyone who has looked 
through a microscope. 

It may be convenient, if one lives in water and doesn't want to go anywhere in 
particular, to have a jellylike consistency, but most animals have some sort of device 
for stiffening the body. This strengthening matter, the skeleton, may be either ex- 
ternal or internal. An external skeleton serves secondarily as a means of protection. 
The most common material of skeletons is calcium, either in the form of carbonate 
or phosphate or as a combination of the two, but it may be silica or hornlike chitin. 
Plants as well as animals have a stiffening material, the woody tissue, which, like 
chitin, is a carbohydrate, called cellulose. Silica, too, is present in some plants, giving 
their cutting power to grasses and sedges. Many of the aquatic plants (algae) secrete 
large amounts of calcium carbonate. Most organisms, in short, possess "hard" parts, 
bones, shells, spicules, or woody material. Were it not for these relatively indestructible 
portions there would be little record of the history of ancient life. "Soft" parts, the 
protoplasmic tissues, flesh, and cartilage, decompose very rapidly through the action 
of bacteria. Hard parts, although they decay, do so slowly; consequently there is a 
chance that they may be preserved as fossils. 

What is the process of becoming a fossil? It is merely preservation, either by the 
checking of decomposition or by the replacement of the hard parts by some relatively 
durable substance. Anything unfavorable to the life of bacteria impedes decay. Very 
dry air, a low temperature, sea or bog water, burial in mud or volcanic ash, an en- 
crustation of pitch, gum, or calcium carbonate, all have a more or less preservative 
effect, so that decomposition is either retarded or entirely prevented. When bacteria 
are entirely excluded not only the hard but the soft parts as well may be preserved 
as in a modern refrigerating plant. The most famous instances of cold storage are 
those of the remains of mammoths and rhinoceroses occasionally found in the frozen 
gravels and ice of Siberia. Another case of remarkable preservation is that of insects 


in amber (Fig. i). While it was a sticky gum exuded from a species of pine, numer- 
ous insects were trapped in it, to be preserved as it hardened. Although amber of 
many ages is known, the most abundant insect-bearing material is found in the 
Oligocene strata on the Prussian shores of the Baltic. 

Suppose an organism to be buried in some substance which retards decay. Con- 
ceivably the sand, clay, or calcareous ooze which surrounds it may become sufficiently 
compacted while the organism retains its original form to hold its shape. Then sub- 
sequent decay of the object will leave a hollow mold (Fig. 3). This may be preserved 
as such, in which case it would itself be a fossil, or percolating waters may eventually 
fill it with calcite, silica, pyrite, mud, or even sand. Examples of this sort, with the 
original outer form preserved as a cast, commonly reproduce only the skeleton, for 
the soft parts decay rapidly even when partially protected. Many plants, whose woody 
tissue decays less rapidly than flesh, are preserved in this way, sand-casts of tree 
trunks, branches, and roots being common in late Paleozoic strata. Many mollusks 
and other animals with hard external skeletons are similarly preserved. 

There is another type of replacement in which decay of the skeleton occurs in 
the presence of mineralized waters, so that for each particle removed the water gives 
up a bit of its mineral matter, producing a delicate replica of the whole original 
structure (Fig. 4). The conditions favorable to this process have, however, rarely 
obtained. A few localities have furnished most such specimens, a noteworthy one 
being a thin layer of coal in the Carboniferous of England. The so-called "coal balls" 
found in it are really calcareous concretions in which a part of the vegetation which 
formed the coal has been replaced by calcium carbonate. Those who have visited the 
Petrified Forest of Arizona or other areas of "bad lands" in the western states have 
probably noticed the great amount of petrified wood which, although entirely changed 
to stone, still shows the characteristic rings of growth, knots, and other features of 
modern trees. This type of preservation is much more commonly found in the 
replacement of plants than of animals. 

Some fossils are so preserved that they retain indications of the shape of the 
internal organs of animals, even though no tissues actually remain. Thus there are 
certain creatures which commonly ingest large quantities of mud with their food. 
The mud-filled alimentary tracts of a few such organisms have been recovered, 
showing the shape of stomach and intestines (Fig. 3, at right). Coprolites, the fossil 
excrement of animals, also show something of the shape of the alimentary canal but 
are particularly interesting because many of them contain undigested remains of food. 
Impressions of skin are sometimes found, and, more rarely, impressions of other soft, 
perishable tissues. Even jellyfish, whose bodies consist mostly of water, and other 
equally delicate organisms have been found at a few places where the rock is of fine 
grain. The two most famous localities for such fossils are the outcrop of Mid- 
Cambrian strata above Burgess Pass, near Field, British Columbia, discovered and 

FIG. i. Contrasts in states of preservation. At left, a natural cast in sand- 
stone of the interior of the shell of a brachiopod. Coarse-grained though the 
matrix is, the heart-shaped muscle scars are conspicuous. At right, a slab of 
sandstone, showing the trail of a dinosaur; halfway along the slab it stepped 
into the water. Between the two, an insect in amber. Photograph at right 
through the courtesy of W. E. Corbin. 

FIG. 2. A cretaceous sponge, seen trom the side and in median-section, 
showing form, canals, and cloacal cavity into which the excurrent canals 
empty. Many sponges are preserved in this way, for particles of sediment 
became entangled in the spicular mesh, permitting the retention of the original 
form. From J. F. Pictet, Traite de paleontologie. 

Fie. 3. At left, a flint nodule from the English chalk, which, on being 
split open, revealed the cavity formerly occupied by a clam shell. From such 
a mold an artificial cast can be made. At right, an unusually well-preserved 
amphibian from the Pennsylvanian with mud-filled cast of the stomach and 
alimentary canal. From R. L. Moodie. 

FIG. 4. At left, a magnified transverse section of a calcified stem of 
dodendron, showing cell structure. At right, a carbonized Pennsylvania!! 
fern. Photograph at the left by courtesy of W. C. Darrah. 

FIG. 5. Dinosaur bones partially uncovered in quarry. Photograph by 
courtesy of Dr. Barnum Brown. 


explored by Dr. C. D. Walcott (Figs. 7, 10), and the quarries in the lithographic 
limestone of Upper Jurassic age at Solenhofen, Bavaria. These latter have been worked 
for many years and have furnished some of the most remarkable and important of 
fossils, running the gamut of the animal kingdom from jellyfish to flying reptiles and 
toothed birds. These localities will be repeatedly mentioned in the succeeding chapters. 
Skeletons which consist essentially of compounds of carbon that is, those of 
plants (Fig. 4), fish, and invertebrates with chitinous coverings, such as insects, crus- 
taceans, and similar animals are commonly preserved as a black, filmy residue, 
showing the form more or less perfectly but much or even entirely flattened. The 
change in this case seems to be a chemical one, involving the loss of the volatile con- 
stituents and reducing the composition to a state approximating that of coal. The 
bark of plants is especially resistant to decay; hence it is not unusual to find rhizomes 
and stems of trees represented by carbonized remains on the surface of a cast of the 
interior woody tissue. Some fossil fish are similarly preserved, the blackened scales 
surrounding a mass of matrix, as if the skin had been stuffed by a clever taxidermist. 
Artificial structures made by organisms are occasionally found. Some burrowing 
animals lined their habitations with bits of shells or sand .which they cemented to- 
gether, whereas others made nests which are occasionally preserved in the rocks. The 
most abundant artificial structures are the various implements of prehistoric man. 
To recapitulate, the states of preservation in which fossils are found m^y be 
tabulated. : 

Actual remains of organisms 

with the soft tissues preserved 
with only the skeleton preserved 

with the skeleton partially changed by addition of mineral matter 
with tissues carbonized 
Natural molds of organisms 
Replacements of the hard parts of organisms 
Skeleton petrified 

without loss of structural details 
with loss of structural details 
Skeleton replaced by natural casts 

mechanically, by infiltration of sand or mud into natural molds 

(often called external casts) 

chemically, by deposition of mineral matter in the same 
Natural casts of cavities in the hard parts (commonly called internal casts) 
Impressions of organisms 
Tracks, trails, and burrows 
Artificial structures made by organisms 


The paleontologist must be familiar with all these various states of preservation, 
for the same sort of animal may appear to him in various guises. It will have one 
appearance as a mold, another as an internal cast (Fig. i), and, as we have shown, 
there are sundry other possibilities. Furthermore, the shells or bones may have been 
crushed or distorted by pressure during their long retention in the rocks. Generally 
they are decolorized, and the original substance may have been replaced by sand, mud, 
silica, opal, carbonate of lime, barite, pyrite, marcasite, collophane, or, more rarely, 
salts of copper or even of silver. 

The term "fossil," which referred originally to anything dug up from the earth, 
has been restricted to the various classes of organic remains listed above. It is diffi- 
cult to frame a brief definition which is both inclusive and exclusive, but it may be 
said that fossils are the remains of organisms or the direct evidences of their former 
existence, preserved by natural causes in the earth's crust. This definition, although 
generally acceptable, has the fault of being rather too inclusive, since it makes no 
reference to the time of burial. As a matter of convenience many paleontologists, 
including the author, arbitrarily exclude from the category of fossils all things which 
have been buried since the beginning of historic times. Such a course helps the 
paleontologist to avoid duplicating the work of the archaeologist, botanist, and sys- 
tematic zoologist, and yet leaves a sufficiently indefinite line of separation to enable 
each paleontologist to decide for himself how nearly he shall approach modern times. 
Care should be taken not to confuse the terms "fossil" and "petrifaction." To petrify 
is to turn to stone, and it is evident from what has been said that not all fossils are 
petrified. On the other hand, not everything which is petrified is a fossil, for to be a 
fossil an object must be organic in origin. So far as possible the adjective "fossilized" 
should be avoided. It is at best a nearly meaningless word, usually employed as a 
synonym for "petrified," with which it is not synonymous; it is therefore much 

Because the prime necessity for the preservation of an organism is that it shall 
be protected from decay by some covering, it follows that aquatic animals and plants 
are much more apt to be preserved than the inhabitants of the land. Consequently, 
marine invertebrates and fish are much more common as fossils than terrestrial 
organisms are. A land animal stands little chance of becoming a fossil unless it 
happens to die in a bog or by a stream, although the preservation of bones in caves 
under the protection of a stalagmitic or earthy cover is common. 

The task of the paleontologist is to reconstruct from such materials as he finds 
in sedimentary rocks the animal and vegetable life of the period during which the 
strata were in the process of accumulation. Too often he must rely upon incomplete, 
distorted, and broken objects which are not easily interpreted. The solution of his 
difficulties can come only through comparison: first, with other materials from the 
same strata, in the endeavor to bring together scattered parts belonging to the same 


sort of animal or plant; and second, with such modern organisms as appear to be 
related. A wide knpwledge of the comparative anatomy of modern organisms is 
therefore a necessary part of the equipment of the paleontologist. Fragmentary material 
must be pieced together to build up a whole skeleton. Then from scars of muscles, 
shapes of bones or shells, and such other features as may be preserved, it may be pos- 
sible to arrive at a rough approximation of the form of the animal as it appeared in life. 
Such restorations, however, are always to be labeled as tentative, for the entire anatomy 
and osteology of few extinct animals is satisfactorily determined. Some things, such 
as coloration, length of fur on mammals, amount of fat, and the like, must be inferred 
almost entirely from our knowledge of living creatures. We cannot prove, for in- 
stance, that any fossil camel had a hump. On the other hand, some soft parts, 
although boneless themselves, may be recorded; for example, the proboscis of an 
elephant is registered by recognizable modifications of the nasal bones of the skull. 

Knowledge of paleontology obviously progresses not in a single, direct line but 
by irregular steps along a wide frontier. New materials are constantly coming to 
light, revealing not only new kinds of animals and plants, never before seen by man, 
but new facts about extinct organisms long imperfectly known. Each so-called species 
of extinct creature is really an artificial creation, built up by man on the basis of re- 
mains found in the rocks. Each represents the sum total of the best information 
available at the time, but every scientist admits that his conception of a species is 
liable to change with the discovery of new material or new methods of study. It fre- 
quently happens that what is called a single species by the original describer will be 
seen as four or five or ten species by some later worker with a wider knowledge of 
the subject. 

If one is to be able to speak of any particular kind of animal or plant, it must have 
a name; so to each kind, or species, a name is given by the person who first publishes 
a description of it. Many experiments were made before a definite system of nomen- 
clature was finally reached, about the year 1758. It was natural to try not only to 
assist the memory by applying a descriptive name but also to indicate the relationship 
of the particular animal or plant to other organisms. Men naturally like to get their 
knowledge into as orderly, usable, and easily remembered form as possible, and so 
with the naming there became involved the idea of classification. It is obvious that 
certain modern animals are more closely related to each other than they are to others. 
Anyone would say at first glance that a cat and a tiger are more closely related to each 
other than either is to a cow. Yet the cow is more like a cat than it is like a fish. 
Thus the classification of plants or animals is built up about degrees of likeness or 
unlikeness. The name given to the particular kind is intended to furnish a knowledge 
of at least one degree of relationship, as well as to serve as a convenience in mention- 
ing it. The earlier writers, who were unnecessarily descriptive, gave names a whole 
sentence long. Linnaeus, the great Swedish naturalist, about 1758 set the fashion now 


followed, of cutting the name down to two words, the first or generic name indicating 
relationship, and the second or specific name suggesting, ideally, some outstanding 
characteristic of the organism described. It is usual to compare the generic name with 
the family name among people and the specific name with the Christian name. The 
generic name is given to a group of species which are found to be very closely re- 
lated in the structure of their bones, teeth, muscles, et cetera, but each of the various 
kinds within a group has a specific name. Thus the genus of the cats is Felts \ the lion 
is Felis leo, the tiger Felts tigris, and the house cat Felts catus. The generic name 
shows their evident close relationship, and the specific name indicates which particular 
kind of cat is meant. 

It will be noted that all of these names are in Latin or latinized Greek. This 
seems to many a great drawback and may make the study of natural history repellent. 
One can hardly blame a beginner for finding such names as Strongylocentrotus 
droebachiensis and Sphaerocoryphe pseudohemicranium rather clumsy at first, but 
few scientific names are as bad as that. The Latin names are, as a matter of fact, a 
tremendous advantage, for they are in general use all over the world. Although one 
may not be able to read any other part of a Russian or Japanese book on natural 
history, one can at least understand what animals and plants the author is enumerating. 
If one had to know the common names of the organisms in all languages, dialects, 
and localities of the world, it would be impossible to get any idea of the fauna and 
flora of the globe. Everyone is familiar with the fact that in various sections even 
of the United States common names of animals and plants are differently applied. 

There are different ideas about the relationships of various organisms to one 
another; hence one should not expect to find textbooks in agreement about classi- 
fication. In a general way, the species is the unit in the system, just as the inch is 
the unit in the widely used British system of mensuration. Although it is true that 
no definite number of species is required to make up a genus, the next higher unit, 
there is nevertheless a feeling that if a genus has a very large number of species it is 
capable of subdivision; hence subgeneric names may be employed to designate groups 
of closely related species. Genera are brought together in families; large families 
may be split by the erection of groups called subfamilies. Families are gathered under 
orders, or in some cases, into superfamilies under the orders. The orders are con- 
sidered as subdivisions of classes or, in large classes, of subclasses. The classes are 
majoi* groups under the phyla, the phylum being the largest unit ordinarily used, 
although the term "kingdom" is still in use for the two great divisions of organisms, 
animal and vegetable. 

Although the species is the unit, it can be subdivided, just as the inch can be 
divided into a certain number of barleycorns and lines. Many systematic zoologists 
recognize subspecies, and even give names to varieties, so that the Latin name may be 
a trinomial or even a polynomial, thus reverting to the condition which obtained 


before the time of Linnaeus. Fortunately, few paleontologists have got beyond the 
trinomial; most of us retain the binomial system. After all, a name is a purely artificial 
thing, employed for convenience. If the endeavor to make it descriptive, either of 
characteristics or relationship, causes it to become so long as to be cumbersome", it 
ceases to fulfill its original purpose. 

A surprising number of students each year ask for an example of the use of these 
terms; one is therefore inserted here. 

Kingdom Animalia. Includes all animals 
Phylum Chordata. All animals with a notochord 
Class Mammalia. Animals which suckle the young 
Subclass Eutheria. Mammals with placenta 
Order Primates. Mammals with flat nails 
Suborder Anthropoidea. Tailless, semi-erect or erect primates 
Family Hominidae. Erect, large-brained anthropoids 

Genus Homo. Anthropoids with the modern type of brain 
Species Homo sapiens. Men with chins, "even as you and I" 

Many an innocent youth will in later life express himself in print. He should 
remember that generic names begin with a capital letter, specific names with a small 
one. Both should be printed in italics. 



Fossiles give joy to Galen s soul, 

He digs for knowledge, like a Mole; 

In shells so learn'd, that all agree 

No fish that swims knows more than he! 

John Gay, "To a Lady" 

e are fossils found ? Many people know them only as exhibited in museums 
ancL^fo not encounter them in the ordinary routine of their lives. In my own case, 
Lfcegan as a schoolboy a collection of the rocks and minerals to be found within ten 
or fifteen miles of my birthplace in southwestern Connecticut and was naturally led 
to read such books on geology as were in my home or in the local library. Since these 
books devoted a great deal of space to fossils, it became my highest ambition to add 
some of them to my collection, and even though the books explicitly stated that fossils 
did not occur in granites and gneisses, the rocks of the surrounding country, I spent 
many a fruitless day searching for them. The books did say that they were to be 
found in limestone, but I hunted through the limestone quarries north of my usual 
haunts with no better success. As a result of years of such efforts I became firmly 
convinced that fossils were scarce objects, an impression which I have since found 
to prevail not only among those who live upon the ancient metamorphic rocks of 
New England but also among those who dwell in regions where the strata are really 
highly fossiliferous. It is extraordinary how unobservant many people are. Some 
years ago I was collecting in northern Alabama, and during the morning I picked 
up a few arrowheads. We had occasion to stop with a local cotton-grower for lunch. 
I asked bim if Indian implements were common thereabouts. He replied that he had 
never sesn any. While waiting for the meal to be prepared, I found a couple of dozen 
very nice points in his driveway and garden. In fact, I became so absorbed in the 
searclythat I unobservantly walked into his beehive, and got properly stung. 

As a matter of fact, fossils occur almost everywhere. There are several localities 
* or em even among the "everlasting hills" of New England. They may be expected 
in an /e gj on o f unmetamorphosed sedimentary strata, and if one looks at a geological 
map ot J or th America, he finds that such rocks cover a much greater area than those 
of igneou or metamorphic origin. There is, it is true, a huge area of the latter ex- 
tending fn^ Hudson's Bay to the Labrador coast, and there are two long narrow 
strips of them on e down the eastern and one down the western side of the continent. 
Much greater a? a s, however, are underlain by fossiliferous sedimentary strata. One 


such is the coastal plain from Massachusetts southward to Mexico; another covers 
the vast region south of the Great Lakes and the St. Lawrence, extending to the 
Gulf of Mexico and westward practically to the Pacific coast; and still another, 
an extension of the last, includes nearly all of Canada west of eastern Mani- 
toba, northward to the Arctic and northwestward through Alaska. In fact, three- 
quarters of the earth's surface has beneath its mantle of soil stratified rocks that are 
more or less fossiliferous. They are present beneath the surface o the ocean as well, al- 
though not easily reached. Mr. Henry C. Stetson has recently collected Cretaceous fos- 
sils from strata in place on the wall of a canyon on the southern side of Georges bank 
in latitude 4024'3o" N. and longitude 6807'3o" W., at depths of from 1600 to 1950 
feet below the surface. Fossils are not, therefore, rare, but occur in inexhaustible nurn 
bers. Consequently they should be better known and understood than they actually are 

Not all unmetamorphosed sedimentary rocks contain fossils, of course, nor are 
they entirely absent from metamorphosed sediments. In general it may be stated tha4 
they are more apt to be present in limestone than in any other kind of rock, and lea^sl 
apt to be found in fine-grained shale or coarse-grained sandstone, particularly it. the 
rocks are red in color. There are, however, exceptions to every rule, and any sedi- 
mentary rock deserves a search. * 

Hunting for, and collecting, invertebrate fossils requires no particular prepara- 
tion or equipment, although the person who knows what he is looking for usually 
finds the best specimens, and a certain amount of skill is required if it is necessary to 
break them from the rocks. The paleontologist is often asked how deep one has to 
dig in order to find fossils. As a matter of fact, one seldom does any digging unless 
a particular layer is found to be so productive that it is worth while to remove other 
layers to get at it. If sedimentary rocks had been left by nature in the position in 
which they were laid down, with the older buried beneath the younger, it would be 
necessary to dig deeply to obtain representatives of the more ancient faunas, but there 
are few strata, except those deposited in the deep oceans, which have not been sub- 
jected at some time to folding, faulting, and uplift, so that rocks of all ages Appear 
at the surface at one place or another. To find their fossils it is only necessary to^Visif 
the exposures in cliffs, along the sides of ravines, in stream beds, quarries, and excava- 
tions along railroads and highways. In some cases one must break the fossils from 
the rocks which contain them, but the best specimens are those which through the 
action of frost, rain, alternations of heat and cold, or other natural causes have been 
freed, or as it is called, weathered, from the matrix. Such processes work most rapidly 
upon the relatively soft strata of the central states from Ohio westward to Minnesota 
and southward to Texas and Alabama. Hence the "interior" region has become 
especially famous for its fine fossils. Collecting under such conditions does not differ 
greatly from picking up shells along the seashore, for really wonderfully preserved 
material can be readily obtained. 


Although invertebrate fossils are common and fully as worthy of study as any, 
the ones which attract most attention are the larger and more showy bones of the 
vertebrates. These are seldom found or collected except by trained men who go to 
the regions where they are known to occur. Occasional specimens are found by people 
not professionally engaged in their study, but these are apt to be ruined by the zealous 
collector, who is always overanxious to secure his specimen and is ignorant of the 
methods by which it might be preserved. It is perhaps worth while to describe briefly 
three of the most famous regions for vertebrate fossils, the chalk of western Kansas, 
the Tertiary deposits east of the Rocky Mountain region, and the dinosaur beds of 
the United States and Canada. 

The chalk deposits of western Kansas were first explored in the years between 
1870 and 1875. At that time the roving bands of Indians that infested the country 
made it necessary for expeditions to be accompanied by detachments of troops from 
the near-by army posts. Indians were not, however, the only or the worst difficulties 
in the way of the searcher for fossils in that region. It was then almost completely 
unsettled, as large parts are even today. Since the chalk beds are bare of vegetation 
except for a few shrubs, the howling winds blow the calcareous dust into eyes, nose, 
and mouth, and cause painful inflammations. Furthermore, in an expanse of country 
a hundred miles long and forty wide, there are only a half dozen springs of fresh 
water, and to make camp the early collectors often had to hunt for hours for moist 
ground in which the borings of crayfishes indicated the presence of water a short 
distance beneath the surface. When such a place was found and a well dug, a supply 
of alkaline water was procured, but, although liquid, it was exceedingly disagreeable 
to the taste and had extremely unpleasant effects. 

The prize fossils of this region consist r r - fish ten to fifteen feet long, aquatic rep- 
tiles fifteen to fifty feet long, flying icptiles, and toothed birds. Although the rock is 
soft, it is not possible to dig out the individual bones of the skeleton separately, for all 
are crushed, distorted, and flattened to such an extent that, if they were separately re- 
moveu, most of the thinner ones would be broken and destroyed, the larger ones hope- 
lessly mixed up, and both so moved from their natural association as to lose much of 
their significance. The material must therefore be taken out in slabs. The finest speci- 
mens are those which, when found, are entirely covered by rock excepting for the tip of 
some extremity such as the snout, tail, fin, or paddle. Since few of the skeletons are 
complete, when one sees a bone, or bones, projecting on an outcrop, the first thing to be 
done is to determine what the animal is and how much of the specimen still remains. 
No bones are removed until the rock has been cleared from the entire surface. Then if 
the exploratory work shows that all or a considerable part of a skeleton is present, and 
not, as too often proves to be the case, merely one or two odd bones, enough of it is 
uncovered to learn its extent. When it has been outlined, and as much of the cover 
has been taken away as seems safe, a trench is dug around the specimen, the depth 


varying with the size of the fossil and the thickness of the layer in which it is em- 
bedded. If the animal is small, covering an area of no more than five by seven or eight 
feet, an attempt is made to get it out in one slab, the exposed surface having first been 
covered with wet paper, or cheesecloth soaked in poisoned gum arabic or glue, to 
make the bones firm and keep them in place. Often it is necessary to follow this 
coating with burlap soaked in plaster of Paris. After the upper surface has been 
secured against breakage, the layer in which the fossil is preserved is pried up by 
means of wedges, picks, and bars, and tilted up with levers; a frame is built around 
it, the specimen is turned over, and the box is completed. Then it may be hoisted 
with an improvised derrick onto a wagon or truck and started for the railroad. When 
specimens are too large to be moved as one slab, they are cut into the requisite number 
of pieces, the division being made where as few bones as possible will be affected, 
care being taken to preserve all parts of broken bones. In some cases an almost 
entire skeleton has been found which has been exposed so long that it is badly 
weathered and seems hardly worth collecting. Such specimens have been saved by 
coating the upper surface with plaster and cleaning the matrix from the lower side, 
a process which reveals the unweathered sides of the bones, the plaster taking the 
place of the original rock. 

The fresh-water Tertiary deposits of the high plains east of the Rockies, and of 
the intermontane valleys within and west of them, provide an extensive field for 
search in beds ranging in age from Paleocene to Pleistocene. Almost all the states 
in the vicinity of the Rocky Mountains and westward to the Pacific coast have fur- 
nished one or many good localities for fossil mammals, but the Oligocene and Mio- 
cene beds of South Dakota, western Kansas, western Nebraska, and eastern Colorado 
and Wyoming have been especially prolific. Modern expeditions to these regions 
maintain a central movable camp from which one or several trained collectors go out 
to spend their days in prospecting along the sides of the bluffs and dry "draws" of 
the "bad lands." Although living conditions are usually not so bad in the Tertiaries 
as in the Cretaceous chalk, the temperature may be extremely high, water bad or 
absent, and surfaces glaringly white in the burning sun. Most mammals are found 
as individual teeth, bones, skulls, or parts of jaws, but once in a while, perhaps two 
or three times in a season, an entire or nearly entire skeleton is found. In reading 
descriptions of new species one frequently finds that the paleontologist has written 
somewhat as follows: "This splendid skeleton was found by Mr. Smith in the lower 
part of the Oligocene near Pole Creek, Nebraska. It lacks only the skull, backbone, 
pelvis, manus, and pes, but the remainder of the specimen is in a marvelous state of 
preservation, except for some crushing and distortion of the limb bones." When 
two or three bones are gathered together, the paleontologist waxes enthusiastic. 

The process of excavating a single bone or a few bones of the skeleton of a mam- 
mal is relatively simple. If only a skull or a few large, solid bones are present, they 


are dug out with pick and awl, care being taken not to remove all the matrix, since 
the extra rock will make shipment less perilous. To extract a whole skeleton or fragile 
bones of any sort, it is necessary, as has been said, to proceed with great care, cutting 
a trench around the specimen, saturating bones and matrix with gum arabic or 
shellac, bandaging them with strips of cheesecloth soaked in poisoned gum or paste, 
and finally reinforcing all with splints held by burlap soaked in plaster of Paris. 

At a few localities bones or skeletons have proved to be so abundant as to make 
collecting a process of quarrying. Such were the deposits at Manhattan, Kansas, in 
the early days and at Agate Springs, Nebraska, more recently. At these places a 
particular layer was found to be made up almost entirely of bones, so that blocks 
were taken out for removal to museum laboratories where the bones could be 
cleaned from the matrix. In these operations, horses, plows, and scrapers sufficed to 
remove the overburden; the layer was mapped out, split into blocks of convenient 
size, the sections numbered, marked to show how they fitted into adjacent blocks, 
and a careful plan drawn so that bones and skeletons broken up in the process of 
removal might be reunited when removed from the matrix. At such places the 
collector becomes a super-quarryman, engaged in a tedious but interesting task. 

The chief localities for remains of dinosaurs in North America are in the third 
region, along the flanks of the Rocky Mountains and the plains to the eastward, from 
Alberta to Colorado, the richest finds of Jurassic dinosaurs having been made in 
Colorado, Utah, and Wyoming, and of Cretaceous species in eastern Wyoming, central 
Montana, and along the Red Deer River in Alberta, Canada. 

These great American reptiles are of comparatively recent discovery, the first 
having been brought to the attention of paleontologists and the scientific public in 
1877, although hunters and travelers had seen the bones and had even brought back 
sections of them as pieces of petrified wood before that. In 1877 three observers, one a 
school teacher, another a professor in the School of Mines at Golden, Colorado, and 
the third a section foreman on the Union Pacific railroad, found dinosaur bones. As 
the late Samuel W. Williston once said, their discovery was not nearly so remarkable 
as that the dinosaurs had remained so long unnoticed. "The beds containing them 
had been studied for years by the geologists of the Hayden and King surveys, yet in 
some areas acres were literally strewn with bones and fragments of bones, and at 
what has since been known as the Bone Cabin quarry in central Wyoming, a Mexican 
sheep herder had built the foundations of a cabin by cording up the huge limb 
bones of dinosaurs." 

One of the discoverers, Professor Arthur Lakes of Golden, was hunting for fossil 
leaves in the basal Cretaceous sandstone near Morrison, Colorado, in March, 1877, 
when he came upon an enormous vertebra partly protruding from the rock. Since 
he had heard of Professor O. C. Marsh of Yale in connection with discoveries of 
toothed birds in the chalk of Kansas, he communicated with him, with the result 


that the great bone, rock and all, was expressed to New Haven at an expense of ten 
cents per pound for transportation. Professor Marsh promptly described, from this 
single vertebra, a new dinosaur which he predicted would be found to have a length 
of 115 feet. Professor Lakes was at once set to work to collect from the beds which 
had furnished the bone, and the name Morrison is still retained for the Upper Jurassic 
formation in which these great reptiles are found. A Mr. O. Lucas, an amateur 
botanist teaching school at Garden Park, Colorado, was another lucky discoverer, 
who, also in March of 1877, stumbled on some bones in a little ravine not far from 
Canyon City. He had heard of Professor E. D. Cope of Philadelphia, who later be- 
came Marsh's rival, and who was one of the most brilliant naturalists this country 
has ever produced. Cope described a new dinosaur from the fragment sent him, 
but one of Marsh's collectors beat him to the locality, which later produced some 
remarkable specimens. The third discoverer was the famous Bill Reed, long noted as 
a professional hunter, section hand, preparator, and, at the time of his death a few 
years ago, curator of paleontology in the University of Wyoming. His discovery was 
made in the vicinity of Como, Wyoming, also in 1877. 

In the wild scramble to dig out bones which the rivals, Marsh and Cope, de- 
scribed and supplied with names, all sorts of valuable and important skeletons and 
bones were hacked to pieces and lost, but later, as the collectors learned the value 
of time, patience, gum arabic, shellac, glue, and plaster, better and better specimens 
were obtained. 

Hunting for dinosaurs is the same process as hunting for any other vertebrate 
fossil. The areas and formations which can be expected to yield specimens are well 
known, and because of the large size of the bones one would think that by this time 
all possible specimens must have been discovered. But anyone who has been in the 
"bad lands" realizes how innumerable are the little dry draws or ravines that inter- 
sect the hills, how great the exposures, and above all, how rapid erosion is when it 
does rain. Instead of being exhausted, each year sees the dinosaur country producing 
new and more nearly perfect specimens, for the storehouse appears to be constantly 
renewing itself. As with other vertebrates, it is much more common to find individual 
bones than entire or even partial skeletons. When a good skeleton is found, the 
pleasure of prospecting is over, and the collector sits down to one, two, or even three 
seasons of hard work. It can readily be imagined that, if one is lucky enough to 
detect the tip of the tail of a seventy-foot dinosaur projecting from a cliff of hard 
sandstone, it will take some digging to get to the skull at the other end of the verte- 
bral column if it chances to be there. The process settles down to a regular quarry- 
ing operation, except that, for fear of shattering the specimen, one has to use much 
less dynamite than he would like to. In most cases the overburden, down to a re- 
spectful distance from the specimen, may be removed by drilling and blasting. Then 
begins a slow process of exploration to locate and trace the skeleton. As each bone 


is uncovered it must be hardened with gum or shellac, pasted, and wrapped, as already 
described (Fig. 5), After the initial removal of the overlying rock, the greater part 
of the work is prosecuted with a shoemaker's awl and a whisk broom, seemingly 
inadequate instruments for an attack upon a dinosaur. As might be expected, various 
ingenious shifts have been employed in getting out the great skulls and bones of the 
dinosaurs. A single vertebra of Diplodocus or Brontosaurus in its petrified state may 
weigh from 120 to 150 pounds, and a limb bone 600 to 800 pounds. Among the hardest 
things to handle are the huge skulls of ceratopsians, which have in some cases been 
brought back in one piece. The largest single fossil so far known to have been 
handled was a skull of Triceratops, which when boxed weighed 6,850 pounds. When 
one thinks of loading a box weighing nearly three and a half tons, and dragging it 
seventy miles over a roadless tract of canyons and gullies, he realizes that the collector 
needs something more than luck to accomplish his task. 

After the specimens have been wrapped, crated, and shipped to the laboratory, the 
most interesting but at the same time the slowest and most monotonous process in the 
business begins. The specimens in their packings are mounted on circular revolving 
tables, where the box, bandages, et cetera, are removed. In modern work this is easy, 
as the gum and plaster readily soak off, especially when the precaution has been taken 
to put a layer or two of paper next to the bone. In the old days of shellac and plaster, 
however, the taking off of the bandages was no simple matter. I have seen the whole 
outer layer of a bone come off with the bandage, which meant infinite labor on the 
part of the preparator in removing the fragments from the bandage and sticking them 
back on the bone. Even with the best collecting, it takes three men two or three years 
in the laboratory to clean up one dinosaurian skeleton, and another year to mount it. 
Oftentimes it is found that the bones are hard on the outside but powdery within, 
which means that a plaster composition, plus an iron rod as a support, must be run 
into the bone. Or the bone may be complete but so shattered that it has to be taken 
to pieces a little at a time, and the fragments cemented back into place. 

In mounting a specimen, the modern principle is to give it a lifelike posture, with 
as little as possible of the support showing. In the case of big dinosaurs, temporary 
mounts of parts of the backbone are made, plaster casts of the undersurface of the 
vertebrate secured, and a steel support cast from the mold so produced. The various 
parts of the axial skeleton are adapted and fastened to one another, then attached to 
a couple of upright rods, usually pipe of suitable diameter. The legs are held in 
place by fitting half -oval pieces of iron to the inner surfaces and bolting or otherwise 
fastening the bones to them. Museum technique is constantly being changed and 
improved as zealous preparators strive to excel one another. Possibly few paleontolo- 
gists realize how much they owe to the skill of the men who actually handle the 
bones and assemble them in the only positions which their shapes and articulations 
show to be possible. 

FIG. 6. At left, four sorts of Foraminifera. The three at the left have 
calcareous shells; the next is of the arenaceous type. From Joseph A. Cushman. 
At right, a Carboniferous radiolarian. From D. Rust. 

FIG. 7. Three Mid-Cambrian fossils from the Walcott quarry at Burgess 
Pass, near Field, British Columbia. At left, a branching sponge, one of the 
oldest known colonial animals. In middle, a sponge showing large siliceous 
spicules. At right, an annelid worm with long bristles on the parapodia. 
From C. D. Walcott. 

FIG. 8. An archaeocyathid sponge. A, a vertical section, showing central 
cavity, inner and outer walls, and pores. From V. Okulitch. B, C, the spicular 
mesh. After R. and W. R. Bedford. B and C are enlarged, A reduced in size. 

FIG. 9. At left, a modern annelid worm, Amphinome. From an original drawing. At right, 
an Eocene encrusting bryozoan, to illustrate the small size of the apertures of the individual 
zooecia. Six times natural size. From Canu and Bassler. 

FIG. 10. Three Mid-Cambrian fossils. At left, a swimming gastropod, 
HyolitheSy with operculum and foot-supports. In middle, a crustacean, Waptia, 
superficially similar to the modern Apus. From C. D. Walcott. At right, 
a crustacean, Sidneyia, with eurypteroidal form. Photograph by courtesy of 
Charles Resser. 



For according to the proverb, the beginning is half the whole business. 


The oldest aggregation of animals and plants which is adequately known is found 
in rocks of Cambrian age. This, the first authentic record of life of the globe, is 
of more than usual interest. What sorts of organisms lived during those ancient 
times, hundreds of millions of years ago? Were they such as exist today, or suffi- 
ciently like them to fit into a classification based on modern animals and plants? 
Do they meet the expectation of believers in the doctrine of evolution in being 
exceedingly simple? 

Before the actual record is examined, it is necessary to review in outline the classi- 
fication of modern animals. For our purposes the animal kingdom may be divided 
into thirteen great groups, or phyla, although this may not be in agreement with the 
best modern practice. Their characteristics will be summarized, beginning with the 

The Protozoa are animals which, although unicellular, have the power of per- 
forming all the important functions of life; namely, feeding, digestion, assimilation, 
and reproduction. The last is accomplished in two ways: asexually, by fission, the 
parent dividing its substance to produce individuals like itself, or sexually, by the 
fusion of two animals and subsequent division. Nearly all Protozoa are exceedingly 
minute. Only two subdivisions include members which secrete shells. The Forami- 
nifera produce in rare instances skeletons as much as an inch or two in diameter, but 
this is exceptional, for most of them are less than a millimeter across. These animals 
have two types of shells, one formed by the agglutinization of foreign objects such 
as particles of silt (Fig. 6), the other a secretion of chitin, of carbonate of lime 
(Fig. 6, at left), or of silica. The agglutinated or "arenaceous" forms appear to be 
primitive. Another group, the Radiolaria (Fig. 6, at right), build tiny siliceous shells 
which have so many openings that they form only the most sketchy framework of a 

The Porifera or sponges are in some respects comparable to the Protozoa, for a 
single specimen is made up of many cells which, although mutually dependent, are 
individually practically like protozoans. This structure suggests that sponges origi- 
nated in aggregations of single-celled animals which did not completely separate at 
the time of reproduction but remained together to form a composite individual, 


Digestion takes place in some of the individual cells, without benefit of a real mouth, 
stomach, or other definitely formed organs (Fig. 13 C). Most sponges have, however, 
a common skeleton, which, like that of the Foraminifera and Radiolaria, is internal. 
This framework may be horny and flexible, as in the common bath sponge, or it 
may be made up of flinty or calcareous needlelike rods, called spicules (Fig. 7) . 

The coelenterates include several sorts of animals, many of them attached to 
foreign bodies and plantlike in their growth. Most of them are colonial; that is, the 
young remain attached to the parents from which they budded. Two of the more 
important groups will be discussed in some detail on later pages. They are the 
simplest animals which are provided with a mouth and digestive cavity, but they lack 
an anal opening, so that waste must be ejected through the mouth. In the single 
cavity are carried on the digestive, reproductive, and excretory functions. Most of 
them show more or less perfect radial symmetry, with the parts repeated in multiples 
of four or six, the mouth being surrounded by a ring of delicate, fingerlike tentacles. 
The most important classes are the corals with calcareous skeletons (Fig. 12 D), 
graptolites (Fig. 15) and other hydroids with chitinous support, and the jellyfish with 

Next in order are the four great phyla which used to be called by the inclusive 
name of "Vermes," or worms. The greatest achievement of this group was the pro- 
duction by its highest representatives, the segmented Annulata (Fig. 9), of a complete 
digestive tract within a body cavity, a head with eyes, an adequate nervous system, 
and an excretory apparatus. Most of the "worms" are bilaterally symmetrical, and 
they are not colonial, but separate individuals with powers of locomotion. Not all 
of them share these characteristics, however, for although some are primitive, many 
are highly specialized parasites. Since all lack skeletons, they are not well known 
as fossils. Two of the groups, the flat worms and the wheel worms, are practically 
unknown in the rocks, and few representatives of the round worms have been found. 
The fourth group, the segmented or annulate worms, have left many more traces, 
such as impressions, burrows, jaws, or tubes. 

The echinoderms, like the coelenterates, have radial symmetry, but they are not 
colonial, and the parts are repeated in multiples of five. Their skeletons, made up of 
calcareous plates set edge to edge, are really internal, although near the surface. This 
phylum contains eight groups, cystoids, edrioasteroids, crinoids, blastoids, sea urchins, 
starfish, brittle stars, and sea cucumbers, all except the last well represented as fossils; 
some of them will be discussed later. 

The bryozoans are minute colonial animals which, although abundant nowadays, 
are seldom noticed. Superficially they resemble coelenterates, having saclike bodies 
with the mouth surrounded by tentacles, but they are more highly organized, having, 
among other things, a complete gut within the body cavity and no radial partitions 
(Figs. 9, 12 C). They secrete lacelike or more solid calcareous skeletons, which are 


common as fossils. They are important to the geologist but of no great evolutionary 
significance, and will not be further discussed. 

The brachiopods, unlike their near relatives, the bryozoans, are not colonial but 
exist as separate individuals. They are perhaps the most frequently met of the fossils 
of the Paleozoic rocks. Although they are still present in the oceans they are rela- 
tively rare and unimportant; hence little attention is paid to them. Like the more 
familiar clam and oyster, the brachiopod has two shells, but they are placed above and 
below the body instead of at the sides. Two somewhat different kinds are known. 
Those known as the Inarticulata (Fig. n) have thin chitinous shells, reinforced with 

FIG. ii. The modern inarticulate brachiopod, Ltugula, showing shells and 
pedicle. From E. S. Morse. 

more or less phosphate and carbonate of lime. The others, the Articulata (Fig. 12, 
A, B), secrete calcareous shells with interlocking processes on the hinge. 

The phylum Mollusca includes three familiar groups, the snails or gastropods, 
the bivalves or pelecypods, and the cephalopods; two less important classes need not 
detain us. Most of them have calcareous shells; the gastropods have a spirally en- 
rolled, unsymmetrical conch, the pelecypods two similar valves joined together at the 
hinge by a ligament, and the cephalopods straight, curved, or spirally enrolled, bi- 
laterally symmetrical, chambered shells. The student should, however, be warned 
that there are numerous exceptions to all three of these general statements. 

The Arthropoda is the largest and most highly diversified of the phyla. The 
members are characterized by bilateral symmetry, a fundamentally chitinous exoskele- 
ton which is periodically shed and renewed, and a segmented body with segmented 
appendages (Fig. 10). Exceptionally, as among the barnacles, a calcareous shell is 


produced. The classes are: the Crustacea (lobsters and the like), with the first pair 
of appendages flexible tactile organs; the Arachnida (spiders, scorpions, et cetera), 
with claws as the first appendages; the Diplopoda (thousand-legged worms), with 
two pairs of appendages on each segment back of the head; the Chilopoda (centi- 
pedes), with poison fangs just back of the head, and the Insecta, the greatest group 
in the whole animal kingdom, with three pairs of walking legs and, typically, two 
pairs of wings. 

The last phylum, the Chordata, includes those animals with an internal axial 
support known as a notochord. In most of the groups the notochord is replaced at 

FIG. 12. A, B, lateral and dorsal views of an articulate brachiopod. C, a 
bryozoan, drawn as though transparent, showing tentacles about the mouth, 
digestive organs, and muscles within the body cavity. After Parker and 
Has well. D, a coral polyp, the upper portion cut vertically to show tentacles, 
mouth, oesophagus, and body cavity with radial mesenteries. Beneath, dotted, 
is the beginning of the skeleton. Redrawn from a wall chart by Paul Pfurt- 

an early period in life by the vertebrae of the backbone; hence the best known of 
these animals are those called Vertebrata. In the latter category are the Pisces (fish), 
Amphibia (frogs and toads), Reptilia (lizards and snakes), Mammalia, and Aves 

Ever since the days of the great Franco-Bohemian paleontologist, Joachim Bar- 
rande, it has been customary to refer to the animals whose remains are found in the 
Cambrian rocks as the first or "primordial" fauna. But the ten thousand or more feet 
of Cambrian strata were not formed in a moment; their deposition is generally sup- 
posed to have occupied from ninety-five to a hundred million years. It is therefore 
absurd to consider all the faunas which succeeded one another during that long time 
as contemporaneous. All geologists recognize three distinctly different faunas in 


the Cambrian, one in the lower beds, another in the middle, and the third in the 
upper part of the strata. It would be just as logical to group together all the animals 
of the Upper Cretaceous, Cenozoic, and Recent as representing the state of evolution 
of life in the late Mesozoic as it is to say that the total Cambrian fauna is the congeries 
which was in existence at the beginning of that period. 

The Lower Cambrian fauna of the world, so far as it is at present described, 
consists of about 455 species, distributed among the phyla as follows: Protozoa, none; 
Porifera, 18.5 per cent; Coelenterata (all jellyfish) and Echinodermata (cystids and 
edrioasteroids), taken together, 2 per cent; "Vermes" (tubes, trails, and burrows), 
4.25 per cent; Brachiopoda, 27.5 per cent; Mollusca (all gastropods, and all but two 
bilaterally symmetrical forms), 11.5 per cent; and Arthropoda (84 per cent trilobites, 
16 per cent other Crustacea), 36.25 per cent. These statistics probably give a more 
accurate idea of the shell-bearing members of the oldest Paleozoic fauna than those 
based upon the Cambrian as a whole. The proportion of sponges and free-swimming 
gastropods is relatively much higher, of arthropods considerably lower, and of brachio- 
pods somewhat lower, than in the lists published by other writers. 

It is evident that the oldest Cambrian fauna is diversified and not so simple, 
perhaps, as the evolutionist would hope to find it. Instead of being composed chiefly of 
protozoans, it contains no representatives of that phylum, but members of seven 
higher groups are present, a fact which shows that the greater part of the major 
differentiation of animals had already taken place in those ancient times. The other 
phyla not represented are the flat worms, wheel worms, round worms, Bryozoa, and 
Chordata, the last the one which contains the most specialized of all animals. It is 
also apparent that the animals living in Cambrian times were not strikingly peculiar, 
since most of them can be assigned readily to phyla erected on the basis of modern 

What, then, are we to conclude? Are we to deny the special creation of the 
modern fauna, only to find that it is descended from animals specially created some 
millions of years ago ? To answer this question, it is necessary to look further at the 
fauna under discussion. Diversified as it is, if analyzed further it proves to be simple 
as compared with that of the present day. 

In the first place, the whole phylum Chordata, from fish to bird, is absent. 

Although the Arthropoda are numerous, making up more than a third of the 
fauna, the only class represented is that of the Crustacea, the simplest of the arthro- 
pods, and 84 per cent of these are trilobites, the most primitive of the crustaceans. 
Very late in the Cambrian a few marine arachnids appeared. 

The Mollusca are all gastropods, the simplest members of the phylum, and nearly 
all are of the most primitive type, their shells being simple, bilaterally symmetrical, 
uncoiled cones (Fig. 10). 

Next in importance to the Crustacea are the brachiopods, which make up 27.5 


per cent of the fauna. Both the Inarticulata and the Articulata are present, but 80 
per cent of the species belong to the former, which is the simpler group. Such of the 
articulates as are found are of the most primitive type. 

The echinoderms are represented only by a few edrioasteroids and cystids, an- 
cestral to all other members of the phylum. 

The "Vermes," likewise poorly represented, probably because of lack of preserva- 
tion, differ from the other animals that have been discussed in that most of the speci- 
mens found belong to the highest phylum, the Annulata, and, moreover, to the most 
highly organized class of the annulates, the Chaetopoda, or bristle worms. 

Few specimens of Coelenterata have been recovered from the Lower Cambrian 
strata, but, curiously, those so far found belong to the Scyphozoa or jellyfish, one 
of the most specialized groups. These animals are absolutely without hard parts and 
are composed chiefly of water, yet Dr. C. D. Walcott found a few impressions of 
them in the Lower Cambrian shales of Vermont. 

Spicules of siliceous sponges have long been known from Lower Cambrian rocks, 
but the oldest complete specimens are of Mid-Cambrian age (Fig. 7). The most 
specialized of modern Porifera are the glass sponges, so called because their frame- 
work is made up of needles of glasslike amorphous silica. The living members of 
this division are classified according to the shapes of the spicules, whether straight 
and simple or arranged at various angles in two planes. It is interesting to note that 
some Lower Cambrian sponges are of the sort with siliceous spicules, and that they 
can be referred to modern orders on the basis of the forms shown by the elements 
of their skeletons. Much more common are the calcareous archaeocyathinids, the 
oldest known sessile organisms (Fig. 8). 

The Protozoa, instead of making up the whole of the oldest known fauna, are 
very poorly represented in it, if they are present at all. A few species of Lower Cam- 
brian Foraminifera have been described from Russia and New Brunswick, but Dr. 
Joseph A. Cushman, the foremost student of this group at the present time, is 
doubtful if any of them really belongs to it. A few poorly preserved radiolarians 
were found in thin sections of rocks collected from the Cambrian of Thuringia, but 
even their describer was doubtful about them. This lack of specimens, however, 
cannot be taken as evidence of the absence of Protozoa from the faunas of this age. 
It must be remembered that the majority of these animals lack skeletons, and that 
all are small and ill adapted for preservation as fossils. The largest of the radiolarians 
are less than one millimeter in diameter; most of the Foraminifera are equally small, 
and many of the latter have exceedingly frail, agglutinated shells which could hardly 
have been preserved in any abundance in rocks millions of years old. 

To complete the survey of primordial life it is necessary to say a word about the 
plants. Up to the present time no traces whatsoever of terrestrial plants have been 
found except in the youngest zone of the Upper Cambrian, which means not only 


that all of the higher types of vegetation but even such lowly things as ferns, mosses, 
and lichens were absent from the most ancient flora. The Algae (seaweeds) are the 
only plants known to have been present^ A few of the most primitive unicellular 
forms happen to be preserved because they secreted calcium carbonate. Walcott 
described other plants, some of which are perhaps red algae, from his famous quarry 
in the Mid-Cambrian at Burgess Pass. As will be indicated later, there is reason to 
believe that diatoms and bacteria were already in existence. 

No one thinks that the Lower Cambrian fauna contained only 455 species. 
Probably some already described have been overlooked in the present survey of the 
literature. Perhaps many are as yet unknown, after seventy-five years of search. The 
present number may be doubled within the next few years as keener observers split 
up the species already described or supplement the list from discoveries at new 
localities. Suppose we add another thousand to accommodate the various soft-bodied 
animals practically incapable of preservation. Even that brings the list to only two 
or three thousand species as compared with two or three million supposed to be in 
existence at the present time. However one looks at the picture, the "primordial" 
fauna was simple and undifferentiated. 

Various writers have been so much impressed by the amount of differentiation 
shown by the Cambrian fauna that there may be a tendency to overemphasize the 
amount of time necessary to produce it. This is perhaps because one is apt to think 
of the great phyla as absolutely distinct from one another, each higher in the scale 
of organization than its predecessor, as shown in the tables in the textbooks. And 
one thinks of each step in advance as having been accomplished only after a long 
period of time. Yet it is a question whether time is a particularly important factor. 
It is natural to think of the phyla as they are represented today rather than as they 
were in Cambrian times. For example, the term "chordata" connotes fish, birds, 
mammals, et cetera, not the lowly backbone-less members of the group, animals so 
simple that some of them were for years supposed to be invertebrates. To estimate the 
real amount of differentiation which had taken place by the beginning of Cambrian 
times it is necessary to compare the most primitive members of the various phyla. 
If this is done, it appears that there were really only three great steps in the progress 
which led from the protozoans to the present great diversity of animals. How long 
it took to organize a unicellular creature we have no idea. 

Starting from the unicellular Protozoa (Fig. 13 A, B) there are obviously two 
possibilities. The new cells produced in reproduction may separate, thus continuing 
to be Protozoa, or they may remain attached to one another, forming multicellular 
animals, or Metazoa. This was a fundamental step in the progress, but there is no 
reason why it should not have occurred as soon as Protozoa began to reproduce. 
Why it happened may be a mystery, but the introduction of the factor of time does 
not facilitate the explanation. 


Once the metazoan stage was reached, there were again two possibilities. The 
members of the colony might retain their individuality, although giving up much of 
their freedom, as in the case of the sponge^ (Fig. 13 C). That this was a successful 
plan of cooperation is shown by the abundance of sponges at the present day, but it 
cannot be called a "great step," for it led nowhere. The other possibility involved the 
uniting of all the cells into one system, with complete cooperation but loss of in- 

FIG. 13. Schematic drawings to illustrate the fundamental steps in the 
differentiation of animals. A, amoeboid protozoan, with no definite shape. 
B, collared protozoan with vibratile cilium. C, a sponge, in which there are 
nonciliated ectodermal cells, and ciliated endodermal ones within the spherical 
cavities. D, section of a simple coelenterate, with mouth, digestive cavity, 
outer ectodermal and inner endodermal cells. The latter differ among them- 
selves in form and function. E, F, transverse and longitudinal sections of a 
primitive coelomate, showing, ec, ectodermal, en, endodermal, and m, meso- 
dermal cells; and c, the body cavity, or coelom. A, B, C, D, redrawn and 
simplified from various sources; E, F, redrawn after Sedgwick and Wilson. 

dividuality. One can imagine various ways in which the cells might have been ar- 
ranged; what actually happened seems to have been the producton of two layers 
forming a hollow sphere. The inner layer is the endoderm, the outer one the ecto- 
derm, and the opening to the central digestive cavity is the mouth (Fig. 13 D). Here 
we have the fundamental characteristics of the Coelenterata. But is this a change 
which requires time? 

Comparing the coelenterates and the sponges, one finds that the primitive mem- 
bers of the former group were, theoretically, freely floating organisms. This belief is 
borne out by the presence of jellyfish in the Cambrian. Sponges, on the other hand, 


were sessile, as shown by attached specimens from the same formation. Freedom 
and progress thus contrast with stability and vegetative growth. 

The coelenterates had acquired a digestive cavity. The next great step was the 
formation of the mesoderm and a body cavity (coelom). This is so profound a change 
in organization that one must admit that it may have required time, though this 
admission really expresses only our ignorance of the connecting links between the 
primitive coelenterates and the primitive coelomates. The transition appears to have 
been from a pelagic organism to one dwelling on the sea floor, from a swimming 
or floating to a crawling mode of existence. The physical change was from a more 
or less spherical form to an elongate one, from spherical to bilateral symmetry (Fig. 13 
E, F). The study of modern animals furnishes some clues to the possible history, but 
since the fossils give no information it will not be helpful to enter into a long 

Once the coelomate stage (body cavity with a digestive tract within it) had been 
reached, all the "great steps" had been taken. Some groups, such as the bryozoans, 
brachiopods, echinoderms, and Mollusca, have gone downhill; only two, the arthro- 
pods and the chordates, have risen above the status of their ancestors. If one dares 
to put a summary of these remarks in the form of a diagram, it might be expressed 
as in the figure below. 


MOLLUSCA-^^"> \^ 




FIG. 14. Diagram to illustrate the relationships of the principal phyla of 
animals, the "worms" being excluded for the sake of simplicity. Their oldest 
representatives, the annelids, represent the most primitive coelomates now 

The inference is that if one should group the animals into super-phyla there 
would be only four, Protozoa, Porifera, Coelenterata, and Coelomata. There seems 
to be no reason why the protozoans, sponges, and coelenterates should not have 
been practically contemporaneous in origin. Although there was a fundamental 
change, it was a simple one. 

The later changes were more complex, but can they be evaluated as to time re- 
quired? The structure of the brachiopod is much more complex than that of the 
bryozoan; yet brachiopods were well on their way at the beginning of the Cambrian, 
whereas bryozoans did not appear till the Ordovician. The primitive echinoderm is 
more "specialized" than the primitive chordate, but the former left skeletons in the 


Lower Cambrian rocks, whereas the latter do not appear in the record till the Upper 
Ordovician. It should, however, always be borne in mind that the earliest repre- 
sentatives of all phyla were probably without skeletons; hence the true chronology 
will probably never be known. 

Glancing back over what has been said of the Lower Cambrian fauna and flora, it 
will be seen that the animals and plants of that time were, after all, simple. Back- 
boned animals and other chordates and all the higher plants were as yet unknown. 
The Arthropoda, Mollusca, Brachiopoda, and Echinodermata, the higher groups of 
invertebrates, were represented by their simplest types only. On the other hand, the 
lower invertebrates, Annulata, Coelenterata, and Porifera, had representatives then 
which were almost as highly organized as any members of the same phyla today. An 
immense amount of evolution has taken place since the Cambrian. In other words, 
we are not driven to belief in an ancient special creation, but to further research on 
the genealogy of organisms. 

Where are we to look for the ancestors of the Cambrian organisms? As was said 
at the beginning of this chapter, the Cambrian contains the oldest assemblage of real 
fossils as yet known. But the Cambrian strata are not the oldest of the water-laid 
rocks. They rest upon older sediments formed at a much earlier period in the earth's 
history. These older rocks are difficult to study, because many in fact, most of 
them have been intruded since their deposition by gigantic masses of hot igneous 
rocks, and most of them have suffered the stresses of one or more periods of mountain- 
building. Geologists estimate that there are from 60,000 to 100,000 feet of ancient, 
pre-Cambrian sedimentary strata, and it is in them that we must look for evidence 
of animals and plants of ages which antedated the Cambrian. Traces of the earlier 
faunas and floras have already been found, and it will be our next task to see what 
has been learned from the numerous searches which have been made for them. 


Observe creation mercifully hidden 
either in an imaginary Eden, 
or buried in some absent-minded spasm 
of a self -genera ted protoplasm. 

Humbert Wolfe, The Uncdestial City 

Ever since Sir William Logan demonstrated that the Cambrian rocks are not 
the oldest sediments but are underlain by vast thicknesses of water-laid strata, geolo- 
gists and paleontologists have been searching for evidences of a truly primordial 
fauna. Some of the pre-Cambrian strata are limestones and shales which appear to 
be little altered in spite of their incomprehensible age; hence they present a constant 
challenge to the investigator. So anxious are geologists to obtain fossils from them 
that anything which remotely resembles an organism is carefully saved and studied 
in the greatest detail. Although many such objects have been described, few have 
been unreservedly accepted as real fossils. No recently discovered "Old Master" is 
subjected to more critical scrutiny or to more acid tests than is a "find" from rocks 
older than the Cambrian. In this chapter only the more important ones, those which 
have been accepted as genuine by eminent "authorities," will be discussed. 

Paleontologists have always before them odds and ends which, although they 
seem to be of organic nature, show no definite structure. They used to be called 
sponges, for lack of a better designation, but are now supposed to be calcareous algae. 
This is due to their resemblance to the "lake balls" or "water biscuits" which at the 
present day are formed in lakes by blue-green algae. These so-called algae are minute 
one-celled plants more nearly related to the bacteria than to the algae. As they re- 
produce they form long filaments and mats of cells which, during their life processes, 
cause the deposition of calcium carbonate from its solution in the water in which 
they live. As each layer of limestone is deposited, it is overgrown by the plants, so 
that in time a concretionary mass is built up. Such cakes are common in the lakes 
of Michigan and of western New York, where they were first noticed by paleontolo- 
gists. When the water biscuits are cut in slices, the sections show a concentric struc- 
ture, with radially arranged, irregular cavities. 

When these objects came to Charles D. Walcott's attention about 1906, he at 
once recalled having seen many similar structures in the Newland limestone of the 
pre-Cambrian Beltian series of Montana. He thereupon collected and studied great 
quantities of the ancient specimens, describing many new genera and species of blue- 


green algae. To one who has not studied them in detail, Walcott's species appear to 
range all the way from those which from their resemblance to the Cambrian Crypto- 
zoon seem to be organic, through doubtful ones to septarian concretions, with side 
lines among the ripple marks and shrinkage cracks, ending with what may be a 
calcareous tufa. None of them shows more structure than a general similarity to a 
water biscuit, although the describer figured what he took to be chains of algal cells. 
These are not convincing, since they are replaced by opaline silica, retain their original 
convexity, and were derived from one of the least organic-appearing of the specimens. 

In the case of the Ordovician and more recent deposits attributed to lime- 
secreting algae, it is possible to identify the plant, at least generically, by the internal 
structure, since the walls of the skeleton are well preserved. That such is not true 
of the pre-Cambrian specimens may, of course, be due to the greater vicissitudes which 
they have suffered. Perhaps the only test which can at present be applied to this 
class of objects is the apparently simple one of whether or not they are of organic 
origin. If they are organic, it is more likely that they are calcareous algae than that 
they are anything else. 

Since the publication of Walcott's paper, geologists have found similar masses 
in the pre-Cambrian strata in the district south of Lake Superior, in the vicinity of 
Hudson Bay, in the Grand Canyon, and in other parts of the world. Although most 
geologists and paleontologists have accepted them as algae, a few, including the 
writer, have maintained a somewhat skeptical attitude toward these "plants." As 
Professor Olaf Holtedahl of Oslo has pointed out, similar concretions have been 
found in situations which preclude the possibility of their having been formed by 
organisms. In any case, there is no reason for applying generic and specific names to 
such indescribable objects. 

Perhaps the most astonishing discovery in pre-Cambrian rocks was that an- 
nounced by Walcott in 1915. Sections made from a specimen of one of the "calcareous 
algae" proved, on examination under high powers of the microscope, to have in 
them minute (0.95-1.3 microns in diameter), somewhat irregular rods which were 
identified as bacteria. Walcott likened them to the modern Micrococcus, but they 
have been more aptly compared by Henry Fairfield Osborn, so far as superficial 
resemblance goes, to some of the nitrifying bacteria which exist in soils. Walcott's 
paper is a brief one, in which he gives credit to Albert Mann of the United States 
Department of Agriculture for the identification, and leaves it to be accepted on 
faith that an organism without hard parts, and less than o.ooi millimeter in diameter, 
would be preserved in identifiable condition from pre-Cambrian times to the present! 

The calcareous algae and the bacteria are the only plants yet reported from the 
Pre-Cambrian. Attention may now be turned to the animals. 

Years ago L. Cayeux described and figured many species of radiolarians, forami- 
nifera, and sponges from metamorphosed quartzites in Brittany. They have been 


widely accepted as pre-Cambrian fossils, although doubts as to their authenticity have 
been expressed. It now appears that the whole controversy is a tempest in a teapot, for 
a French geologist has shown that the strata in question are rtot pre-Cambrian but 
probably Devonian in age. Whatever the age of the rocks, they are so changed from 
their original condition that it is doubtful if the so-called fossils in them are actually 
the remains of organisms. The objects which Cayeux described are almost the 
only European pre-Cambrian remains which have been supposed to be of animal 
origin. Most of the discoveries have been made in America. More fortunate than 
Cayeux, Walcott found what are considered to be real sponge spicules in the Chuar 
of the Grand Canyon. No figures of them have been published. 

Ati^pT^ania lawsoni, described by Walcott, was widely heralded at the time of 
its discovery (1911) as the oldest fossil, since it was obtained from a limestone of the 
Steep Rock series (Huronian), west of Port Arthur, Ontario. The best specimens 
are silicified in a calcareous matrix and have a cylindrical or pear-shaped form from 
one to fifteen inches in diameter. They are described as having inner and outer walls, 
the space between being filled with radial pillars which are laterally connected to 
produce a concentric structure. Walcott compared them to Cambrian sponges, par- 
ticularly to a South Australian genus, which has an inner and an outer wall connected 
by radially arranged tubes. In later papers he has referred to them as "spongoids." 

The Permian dolomites at Sunderland, England, contain specimens which are 
similar to Beltian calcareous algae and to Atifyfynia lawsoni, but they have been 
shown by G. Abbott and by Holtedahl to be of inorganic origin, for they are con- 
cretionary structures resulting from the replacement of limestone by dolomite. The 
process began along joints and cracks in the rock. If the "fossil" was produced that 
way in the Permian deposit, it is likely that the specimens in the Pre-Cambrian were 
also of inorganic origin. 

Walcott described another "fossil" from the Canadian locality under the name of 
Atif(of(ania irregularis. Some excellent specimens of this form have been investi- 
gated by the writer. Thin sections show that it is composed of aggregates of quartz 
crystals embedded in a matrix of limestone. It is, therefore, of purely inorganic origin. 
Similar crystals occur in pre-Cambrian limestone on one of the shoulders of Mount 
Edith Cavell in the Canadian Rockies. At that locality they appear to have been 
formed by silica-bearing solutions which have passed through the limestone. 

The pre-Cambrian strata in Montana which have been so productive of "cal- 
careous algae" have long been searched for other evidences of ancient life. Several 
discoveries were reported from this region by Walcott, who described various trails 
and burrows, ascribing their origin, probably correctly, to the agency of annelids. 
More noteworthy, however, are the numerous specimens which the same writer named 
Beltina danai. This species was founded upon numerous fragments which were 
believed to belong to an arthropod allied to Pterygotus or Eurypterus (see Chapter 


VII). The supposed test is extremely thin and, in most cases, without any definite 
outline. A few fragments, selected from thousands, do remotely resemble parts of 
eurypterids. This may be said of four of the specimens figured by Walcott. Not 
only is the absence of outline an objection to the reference of these specimens to 
arthropods, but an even more significant circumstance is their total lack of surface 
marking. This excludes them completely from the Eurypterida, for even small pieces 
of the tests of these animals show a characteristic series of scales. The fragments are 
probably of organic -nature, however, and they are widespread in Montana and 
British Columbia. It may be that they are of vegetable origin, perhaps remains 
of brown algae, though nothing definite can be ascertained from their structure. 
It seems likely that the "arthropod'* recently described by Sir Edgeworth David 
and R. J. Tillyard from the Pre-Cambrian of Australia is of similar nature, if organic 
at all. 

The alarms rung by the discoveries of pre-Cambrian fossils have been sounded 
so often without real cause that the paleontologist is perhaps becoming unduly skep- 
tical. From the standpoint of one who has seen all sorts of discoveries accepted with- 
out question, it is refreshing to note the inception of a more critical attitude. Paleon- 
tology, like many another science, has suffered from the general human tendency to 
play follow the leader. Paleontologists have been too prone to accept blindly the dicta 
of "authorities." At one time everybody was describing as fossils the inorganic 
Eozoons, just as for the past twenty ye?rs everyone has been finding calcareous algae. 
As a matter of fact, the flora and fauna of the Pre-Cambrian, so far as they are recorded 
by actual fossils, cannot be said, even by the most credulous, to represent more than 
four groups: blue-green algae, brown algae, sponges, and annelid worms. If paleon- 
tologists were called upon for strictly scientific evidence, they could not prove that 
their determination of any one of these groups is correct. 

It is perhaps possible to get a more just idea of pre-Cambrian organisms if we 
make a brief survey of the state of evolution of life during Cambrian times. So far 
as plants are concerned, the evidence is but little more satisfactory than that regarding 
the similar fossils of the Pre-Cambrian. Cryptozoon is common in the Upper Cam- 
brian, but whether the specimens are the secretions of blue-green algae, as is now the 
popular opinion, or of hydrozoans, as was formerly supposed, no man can say. All 
agree that they are probably of organic origin. Other Cambrian remains supposed 
to be of vegetable origin are in unsatisfactory states of preservation, although some 
of those which Walcott described from the Middle Cambrian may be red and brown 
algae. Fortunately, the Cambrian fauna is much better known. As indicated in the 
preceding chapter, trilobites and other crustaceans, gastropods, brachiopods, annelid 
worms, echinoderms, coelenterates, and sponges have been found in Lower Cambrian 
strata. From the state of evolution of each of these groups, a reasonable inference as 
to their antiquity may be made. 


The fact that the Protozoa are unknown as fossils in the oldest fauna is no indi- 
cation that Foraminifera and Radiolaria were not abundant at that time. The animals 
may have been naked; at best, their minute skeletons are ill adapted for preservation 
or recovery. More informative are the remains of the sponges. Although they com- 
prise only slightly more than one per cent of the fauna, Walcott has shown that all 
orders of the siliceous sponges were represented among the fossils of the Mid-Cam- 
brian; hence, we may conclude that their history began far back in the Pre- 
Cambrian. We are not so sure about the coelenterates. Walcott's Mid-Cambrian 
jellyfish is probably authentic; hydrozoans of some sort appear in the Mid-Cambrian, 
and graptolites in the Upper, whereas the corals are represented by a single non- 
calcareous species in the Mid-Cambrian. Differentiation in this group probably took 
place during the Cambrian, although there must have been an ancestor in pre- 
Cambrian times. Since this was written, a jellyfish has been found in the Pre-Cam- 
brian of the Grand Canyon. It has not yet been described. 

The presence in Walcott's famous quarry of numerous specimens which seem 
referable to chaetopod annelids necessitates the postulation of an early pre-Cambrian 
origin for this group. The Cambrian representatives of the echinoderms are all simple 
cystids or edrioasteroids, a fact which indicates that this phylum had not existed long. 
Despite their numerous genera and species, the brachiopods were not in a high state 
of evolution in Cambrian times. The inarticulates were old, but the articulates prob- 
ably made their entrance upon the world's stage about the beginning of the Cambrian. 
There is no reason to suppose that the Mollusca, known chiefly from numerous free- 
swimming gastropods (hyolithids) and a few simply coiled snails, had any long pre- 
Cambrian history, for their Cambrian representatives are surely primitive. The trilo- 
bites and other Cambrian arthropods suggest a different story. Undoubtedly animals 
of this sort had been in existence for millions of years before the time of their first 
actual appearance as fossils in early Cambrian strata. 

Summarizing this brief analysis of the Cambrian animals, it may be inferred 
that the pre-Cambrian fauna consisted of naked Protozoa, siliceous sponges, primi- 
tive coelenterates, annelid worms, inarticulate brachiopods, and trilobites, the latter 
accompanied, perhaps, by creatures resembling their ancestors. 

If this inference be sound, why are there so few pre-Cambrian fossils ? 

Numerous and varied are the answers to this question; six will be considered 

1. Pre-Cambrian fossils were destroyed during the changes which took place in 
the metamorphism of the rocks. 

2. Daly's theory: pre-Cambrian marine organisms had no calcareous skeletons 
because of insufficient calcium in the oceanic waters. 

3. Lane's theory: the pre-Cambrian oceans were acid; this condition prevented 
the formation of calcareous skeletons. 


4. Walcott's theory: all the pre-Cambrian strata now accessible were deposited 
on land in fresh water of low calcium content. 

5. Brooks's theory: pre-Cambrian organisms lacked hard parts because they 
lived in the surface waters of the oceans, where skeletons were detrimental because 
of their weight. 

6. The writer's modifications of the Brooks theory: skeletons appeared after 
the Pre-Cambrian as a result of the adoption of a sessile or sluggish mode of existence. 

Taking these up in order, it may be said that the first explanation holds for most 
pre-Cambrian strata. It is the exception, rather than the rule, to find sediments of 
that age which have not been so completely altered as to change the original sand- 
stone, shale, clay, or limestone into gneiss, schist, slate, or marble. The recrystalliza- 
tion which accompanied these changes generally destroyed any organic remains. It 
is practically useless to search for fossils in metamorphosed rocks, although they are 
sometimes found. There are a few formations, such as the Beltian of Montana, 
the Keweenawan of Michigan, parts of the Huronian of Ontario, and other strata 
in Texas, Newfoundland, and China, which appear to have partly escaped the 
processes of alteration. Their lack of fossils must be explained on other grounds. 

The second theory, proposed some thirty years ago by R. A. Daly, has found 
considerable acceptance. Obviously, soft-bodied animals are ill adapted for preserva- 
tion as fossils, and no calcareous skeletons can be formed if the water in which the 
organisms live does not contain sufficient available calcium. The oceans of the present 
day contain a larger percentage of the bicarbonate of this metal in solution than do 
rivers and lakes upon the land; hence marine organisms build thicker skeletons. 
Moreover, there is a direct relationship between the thickness of shells of Mollusca 
and the concentration of calcium salts in solution in rivers and lakes. For example, 
the fresh-water shells of New England, where there is little calcium, are thin as com- 
pared with those in the Mississippi basin, an area where much limestone is dissolved. 
Daly noted these facts and found a reason for the lack of calcium in the oceans of 
the Pre-Cambrian. In brief, he argues that in the absence of an effective scavenging 
system on the floor of the deep oceans there must have been, over two-thirds of the 
globe, an accumulation of organic matter, which, on being decomposed by bacteria, 
produced much ammonia. This, in his opinion, caused the precipitation of nearly 
al! of the calcium in the oceans and the formation of calcareous ooze, an inert com- 
pound, unavailable to animals. The only calcium in solution thereafter was the 
small amount contributed by the rivers, and itself on the way to being precipitated 
by the same process. It was not until Ordovician times, when scavengers became 
numerous, that much skeletal material was accessible to animals. 

It would be a logical deduction from Professor Daly's theory to suppose that the 
pre-Cambrian carbonate deposits were of deep-ocean origin. So far as I know, there 
is nothing about the rocks themselves to indicate this; Daly himself agrees that there 


is much evidence that they were formed in shallow epeiric seas like those of the 
Paleozoic. Under such conditions, the dissolved calcium bicarbonate brought in by 
the rivers must have passed through shallow seas on its way to regions cold enough 
to cause the water containing it to sink and creep equatorward along the bottom, 
where alone it could come in contact with decaying organic matter. Such a circula- 
tion would necessarily be slow, and inhabitants of the shallow seas would have first 
chance at such calcium as there was. Furthermore, in James Bay and the eastern 
Baltic the waters are of low salinity, yet the Mollusca build shells, as do those of the 
marginal portions of the Black Sea, down to a depth of fifty or more fathoms. Animals 
form shells even in rivers of extremely low calcium content; if skeleton-secreting 
cells are present in an organism, they manage to find the needed material in any 
medium except aqua pur a. 

The third theory, A. C. Lane's, cannot be discussed adequately without wander- 
ing far afield into the realms of chemistry. There are good reasons for believing 
that during the early history of the oceans their waters held so much chlorine and 
other dissolved and ionized chemicals as to make them acid and thus effectively pre- 
vent the formation of calcareous shells. Hence the first skeletons may have been 
composed of chitin or silica. It is probable that this condition did exist in the early 
days of the Pre-Cambrian, for such primitive animals as radiolarians and the oldest 
sponges have siliceous skeletons, and the basis of those of the inarticulate brachiopods 
and of the trilobites is chitin. On the other hand, immense quantities of limestone 
were deposited during later pre-Cambrian times, a happening which would have 
been impossible in an acid sea. 

The man who probably expended more time and energy than any other indi- 
vidual in trying to find pre-Cambrian fossils was Walcott. As a result of his dis- 
couraging experience, over a period of eighteen years, he came to the conclusion that 
almost all the accessible strata older than the Cambrian were accumulated on land 
as fluviatile or lacustrine deposits. The Beltian, he thought, might be partly marine, 
for he firmly believed in the crustacean nature of Beltina danai. He pointed out that 
in Beltian times the continents were larger than at present and that all the deposits 
in which fossils have been sought were formed within their margins. The shallow- 
water origin of the strata is indicated by red beds, ripple marks, and surfaces checked 
with shrinkage cracks. Such shallow-water characteristics are, however, as often found 
in sediments of marine as of fresh-water origin. Furthermore, the bodies of water 
in which the strata were laid down were too large, too permanent, and too similar 
in pattern to the later marine epeiric seas to present much evidence that they were 

W. K. Brooks was one of the first to suggest a theory to account for the lack 
of fossils in the Pre-Cambrian. He believed, with many other zoologists, that the 
superficial waters of the open ocean were probably the place of the first great ex- 


pansion of life, if not the region of its inception. He developed the idea that the 
pre-Cambrian animals lived exclusively in this zone, where they maintained a free- 
swimming or floating existence in which a heavy calcareous shell would have been 
an encumbrance, not a help. Such animals eventually came to live on the sea floor, 
near shore, where the conditions of a new habitat caused them to secrete calcareous 
shells. As it was originally stated this theory had various defects, one of the most 
obvious of which is, as Daly pointed out, lack of reason for the sudden discovery of 
the ocean bottom. Since the time of its promulgation, much has been learned about 
the secretion of shells, so the writer has somewhat modified his explanation, adhering, 
however, to his central idea of the importance of activity. 

That a definite relationship between activity and skeletal armor exists is obvious. 
The common symbols of sluggishness are the snail and the tortoise; such animals 
as the corals and the bryozoans, fixed in one spot, show a maximum of calcareous 
shell and a minimum of flesh. It is commonly said that armored animals are sluggish 
because well protected, and that the unarmored are active because they must seek 
safety in flight. As a matter of fact, cause and effect are reversed in this oft-repeated 
remark; the truth is that animals are probably armored because of their sluggish- 
ness or entire lack of movement. Calcium is present in greater or lesser quantities 
in all water and in many kinds of food. It enters the alimentary tracts of animals 
in solution, but within them it is converted into solid calcium carbonate, either by 
the action of ammonia produced by putrefactive bacteria or by the effect of the 
nascent methane formed during the digestion of the cellulose of plants. The calcium 
carbonate so produced is harmless, but small amounts, still in solution, get into the 
body fluids and reach the protoplasmic cells. Some of these cells apparently pass it on 
to the excretory system; others cause its precipitation in situ and so build a skeleton. 
All animals and many plants are confronted with the necessity of getting rid of 
surplus calcium. Those most active best solve the problem. The production of a 
calcareous skeleton is an involuntary chemical function which will take place in 
animals in any environment. In one sense it may be thought of as a pathologic 
condition, brought on by inactivity. 

As a check, let us examine the Cambrian fauna for connections between activity 
and skeleton. 

Knowledge of the Cambrian Protozoa is too unsatisfactory to allow any profitable 
discussion. The sponges were sessile, and those which survived secreted siliceous 
spicules. The extinct archaeocyathinid sponges were sessile and formed a calcareous 
skeleton. The jellyfish and most worms were active and had none, but the more 
sedentary tubicolous worms early began the formation of calcareous tubes. Cystoids 
and articulate brachiopods were anchored to the bottom, and both secreted cal- 
careous shells, but the wriggling and burrowing inarticulates had only a chitinous 
covering, slightly impregnated with phosphate of lime. Most of the gastropods of 


the Cambrian were free-swimming, or at least moderately active, and formed thin 
calcareous shells. The Crustacea were active and had skeletons which were funda- 
mentally chitinous. In a general way, the rule holds. 

An indication that sessile life was a novelty, newly discovered in Cambrian times, 
is the absence of colonial animals from the Lower Cambrian fauna. Until Mid- 
Cambrian times reproduction appears to have been by fission, or by a truly sexual 
process, for all earlier creatures are separate individuals. The first examples of bud- 
ding reproduction, resulting in the formation of colonies, appear among the sponges 
and hydrozoans found in strata of Mid-Cambrian age. There are but few of these. 
Budding was doubtless the result of the late-formed habit of fixation. Colonial life 
did not become popular until Mid-Ordovician times. 

We can only speculate as to the origin of sessile life. If we turn to the list of 
pre-Cambrian animals, we find that all of them were motile and that predaceous 
carnivores were entirely absent. This group appears first with the few cephalopods of 
the late Cambrian, having missed the Pre-Cambrian completely. In the latter era, 
untroubled by predaceous animals, the swimming and floating organisms must have 
increased rapidly until there came a time when the upper, sunlit parts of the seas and 
oceans were overpopulated. This must have forced some individuals to the bottom. 
It may also be, as E. W. MacBride has suggested, that some of the more sluggish 
animals would naturally drop to the bottom from time to time, simply because they 
were too lazy to keep afloat. Those which fell into the dark abyss of the deep oceans 
mostly perished, for food is scarce outside the zone of sunlight, where alone plants 
can live. Active animals reaching the bottom in shallow water continued to swim, 
or learned to crawl about after food. The more passive adhered to the substratum, 
became relatively inactive, and began the secretion of skeletons because they were 
no longer able to rid themselves of calcium carbonate. 

That the pre-Cambrian animals were all motile seems, therefore, to explain their 
lack of hard parts, and hence to solve the question why so few pre-Cambrian fossils 
are found. That more than are now known will eventually be discovered is probable, 
but it may be predicted that they will prove to have either no skeletons or else thin 
ones composed of silica or chitin. 

It may be well to amplify the discussion of the facts on which what has already 
become known as the "sessile" theory of the origin of calcareous skeletons is based. 
Perhaps the sessile part of the preceding modification of the Brooks hypothesis 
has been overemphasized. The matter is essentially one of degree of activity in rela- 
tion to the elimination of calcium salts, and the hare and the tortoise are perhaps 
better examples than the motile jellyfish and the fixed coral. One must also take 
into account the relative mobility of various parts of the body. It has been suggested 
that if deposition of calcium carbonate is due to inactivity vertebrates should have no 
skeletons, since they are active animals. This is a fair inference but not an argument 


against the idea, for, as will be shown in later chapters, the ancestors of the vertebrates 
had no calcareous internal skeletons. In fact, they had no internal skeletons at all, 
although many sluggish benthonic members of the group did develop calcareous 
external ones. They were succeeded by fish with cartilaginous internal skeletons, a 
primitive condition which has remained unaltered in some to the present day. The 
skeletons of others have been partially calcified. But where? Chiefly in the axial 
region, the least mobile part of a fish. It may be objected that the fish swims by means 
of rhythmic contractions of the muscles of opposite sides of the body, and that the 
axial region is not inert and rigid. Perfectly true. But everyone who has read the 
most elementary books on physics must be aware of the phenomenon of the trans- 
mission of vibratory waves along a wire on which there is interference with free 
movement: the nodes of no motion, and the internodes of movement. It is at the 
motionless nodes in the cartilaginous axial support that calcifications, vertebrae, are 
produced, whereas the internodes remain unossified. The calcareous dorsal and ven- 
tral spines and ribs likewise follow lines of relatively little motion. This is not the 
place to enter into a detailed discussion of the skeleton. Suffice it to say that the head, 
primarily a cartilaginous covering of the brain, is immobile, except for the vertical 
movements necessitated by the opening and closing of the mouth, and that the 
originally cartilaginous portions of this region ultimately became ossified in the higher 
fish. The appendicular system (limbs) of the land-living vertebrates was pre-formed 
in cartilage in the ancestral fish. Here again the segmented nature of the appendages 
indicates a nodal and internodal arrangement, regions of no motion alternating with 
those of mobility. 

So much for one of the major objections to the "sessile" theory or, as the writer 
would have preferred to call it if he had considered it a new theory rather than a 
modification of the Brooks hypothesis, the "sluggard" theory. It might be well to 
state in passing that the slug, although one of the most active of the gastropods, is 
so called because he is, relatively, a sluggard. The term far antedates the naming of 
the animal. 

Perhaps one or two other objections may be forestalled by a few further remarks. 
For instance, it may be pointed out that the sea anemones and stony corals have 
the same structure; both are corals and both are sessile, yet the former have no 
skeletons at all, the latter massive ones. A part, at least, of the explanation of the 
anomalous condition is to be seen in the fact that most of the sea anemones are in- 
habitants of shallow or deep cold waters, whereas most corals are tropical or sub- 
tropical. The matter of temperature as a factor in the formation of skeletons has been 
mentioned above and will be again in the discussion of the Ordovician fauna. It may 
seem that a more acceptable theory could have been built up on the basis of the in- 
creasing warmth of the seas after the Pre-Cambrian. Really thick calcareous skeletons 
do not appear until Ordovician times, the first period in which broad, shallow, warm, 


cpeiric seas were widespread. For example, numerous sorts of arthropods are living 
today, but only one nonparasitic group, the barnacles, contains sessile creatures. And 
the barnacles are the only arthropods with really calcareous skeletons. But, as is well 
known, barnacles are not by any means confined to warm seas. Anyone who has 
tried to bathe off the rockbound coasts of New England is well aware of the associa- 
tion of barnacles and cold water; I think it was Irvin S. Cobb who once said that he 
would as soon have his foot bitten off by a shark as stick it in the water north of 
Cape Cod. 

Various lines of evidence indicate that when the oceans were first fit for organic 
occupancy they were somewhat acid, and that the first plant and animal skeletons 
were siliceous or chitinous. Somewhat later, when animals had progressed as far as 
the stage of the primitive brachiopod and trilobite, the waters had become neutral 
or slightly alkaline but were still "fresh"; that is, soluble salts were present in parts 
per million rather than in parts per thousand. In the course of time calcium bicarbon- 
ate has become more abundant in marine waters, and animals have been confronted 
by an increasingly difficult physiological situation. 

It is possible in fact, probable that animals discovered the sea floor from 
time to time during the Pre-Cambrian. It is unlikely that it would have taken a 
billion years to overcrowd the upper zones of water in the oceans. Even if the rate 
of reproduction were slow, it would have taken only a short time to fill the region 
permeable by sunlight with organisms. But there is evidence that adaptation to 
benthonic life was a slow process. As has been said, sessile and crawling life started, 
so far as the record tells us, at the beginning of the Cambrian, but it was not till the 
Mid-Ordovician, nearly a hundred million years later, that real sluggards became 
common. This hundred million years was probably a quiet period, as compared 
with any previous hundred million. The unstable submarine floor may have been 
repeatedly colonized, but it was probably disturbed by earth movements before the 
lime-secreting habit had been acquired. 

Pre-Cambrian times were the real "dark ages." Many industrious and brilliant 
geologists have devoted their lives to the study of the rocks then formed, but, though 
they have learned a great deal about them, they have been unable to establish a satis- 
factory chronology. Authentic history begins with the oldest really fossiliferous 


Rich with the spoils of Nature. 

Sir Thomas Browne, Religio Medici 

The fauna of the Ordovician differs from that of the Cambrian in its greater 
diversity, many more classes of animals being present. The new groups appearing 
at this time are largely of kinds which secrete thick calcareous skeletons, shells more 
apt to be preserved than the thin chitinous tests of the Cambrian animals. Fossils 
are rare in many Cambrian rocks, but some Ordovician limestones are literally made 
up of them. If one happens to be in the neighborhood of Cincinnati, Ohio; Rich- 
mond, Indiana; Lexington, Kentucky; Nashville, Tennessee; Minneapolis, Minne- 
sota; or other places that could be mentioned, it is worth while to take a trip into 
some of the old quarries on the hillsides or walk along the natural exposures beside 
the streams. Even the experienced collector is amazed by the countless millions of 
3rganic remains which make up the greater part of the Ordovician strata. At most 
si the localities mentioned, the rocks are soft and break down readily on exposure 
to the atmosphere. The fossils, being harder than the enclosing matrix, weather out, 
and one can find specimens as perfect as those on a modern beach. To a person who 
tias a natural bent in that direction there is probably no pleasure more keen than that 
:>f wandering among these debris of ancient ocean floors, picking up the shells of 
mimals which lived countless centuries ago, never knowing what treasure the next 
>tep will reveal. 

The most important of the new arrivals in Ordovician faunas may be enumerated. 
f\mong the coelenterates are the colonial organisms, graptolites and corals. A few 
jpecies of graptolites have been described from the Upper Cambrian, and a little 
evidence for the existence of corals during the Cambrian has been found, but both 
ivere virtually newcomers. The echinoderms increased greatly in variety, starfish, 
:rinoids, blastoids, and sea urchins being represented. Bryozoans came upon the 
icene in profusion, and their numerous coral-like skeletons are highly prized by geolo- 
gists for their value in correlation. Many kinds of gastropods augmented the mollus- 
:an population, and pelecypods, cephalopods, and chitons made their first appearance. 
Sfew groups among the arthropods were the bivalved water fleas, or ostracods, the 
>urypterids, and the sessile barnacles. Remains of fishlike animals have been found 
n the Upper Ordovician strata at Canyon City, Colorado, in the Black Hills, and 
n Wyoming, but the specimens are all badly preserved. Nevertheless, thin sections 


show that these fragmentary fossils really represent early Chordata. By the end of 
the Ordovician nearly all the classes of invertebrate animals except the insects and 
some other specialized arthropods had appeared, but the flora had as yet made little 
advance upon the lowly position which it occupied in Cambrian times. The present 
data, however, are incomplete, for no fresh-water or terrestrial deposits older than the 
late Silurian have as yet been discovered. 

The Ordovician was the time of the first abundant secretion of carbonate of lime 
by animals. This may have been due to one of two causes. It may be that the water 
of the original oceans of the globe was comparatively fresh that is, that it contained 
little dissolved mineral matter. As time went on, the breaking down of the rocks 
on land under the influence of weathering would cause great quantities of soluble 
material to be carried to the sea. It is known that the amount of sodium chloride 
(common table salt) in the sea has constantly increased throughout the ages; it 
may be that the amount of dissolved calcium has likewise tended to accumulate. The 
other possible cause for the increased precipitation of calcium carbonate by animals 
in Ordovician times may have been a change in climate from cold to warm. Water 
containing Carbon dioxide can retain in solution much more calcium than pure water 
can, and cold water absorbs and retains much more carbon dioxide than warm water 
does. Animals living in cold water, consequently, are much less bothered by cal- 
cium than those living in warm water and as a result secrete thinner shells. It is 
possible that the general temperature of the oceans and seas was higher in Ordovician 
than in Cambrian times. Certain it is that, taking the world as a whole, there is 
much more limestone in the Ordovician than in the Cambrian series of rocks. More- 
over, it may be noted in accordance with the above principle that the chief places 
for the formation of limestone at the present time are in tropical and subtropical 
regions. During Ordovician times limestone was formed both under the equator 
and in northern Greenland within a few degrees of the pole, a fact that indicates 
uniformly warm conditions throughout the world. (The reason for the equable 
climate of the whole earth during the greater part of the Ordovician may be seen 
on consulting a palaeogeographic map on which are delineated the lands and seas 
of the time. It was a period when the lands were of low relief and when there was 
a widespread flooding of the continents; the whole extra-polar world enjoyed a 
warm oceanic climate.) Hence it is not surprising that sluggish animals surrounded 
themselves with layers of the easily precipitated mineral. This point is stressed, de- 
spite some repetition, because it is too often stated that animals build hard parts to 
protect themselves. As was said in the previous chapter, the formation of external 
skeletons began in Cambrian times before there were any predaceous carnivores, 
and consequently before there was any need for protection. In the Ordovician, shells 
did serve as armor, and such animals as had them were endowed with a natural 
advantage over others. Hence one may note the operation of both the Lamarckian 


and Darwinian evolutionary processes. The environment (water with dissolved 
calcium) forced shells on the animals. These served, when the time came, as a 
means of protection, and the animals best protected survived. A good example of 
this is seen in the great multiplication of the calcareous-shelled brachiopods, which 
completely overshadowed the chitinous inarticulates in the Ordovician. 

Among the most important animals of the Ordovician seas were the graptolites. 
To understand this group of fossils it is necessary to dwell for a moment upon the 
lowly modern animals known as hydroids. These are numerous both in salt and 
fresh water at the present day but attract little attention, since, if seen, they are 
usually supposed by the casual observer to be small aquatic plants. Their small 
tubular bodies are radially symmetrical, having at the upper end a mouth surrounded 
by slender, fingerlike tentacles. It opens into a simple undivided digestive cavity. 
They are, therefore, coelenterates, closely allied to the corals. They differ from the 
latter, however, in lacking radial partitions within the body cavity and in having a 
life history in which a sessile, plantlike, asexual generation alternates with a free- 
swimming or floating one. In the fixed generation a slender, shrublike form is pro- 
duced by successive budding. This asexual colony eventually produces buds which, 
instead of forming ordinary feeding polyps, become little jellyfish that break off 
from the parent colony and float away. The jellyfish produce the sexual elements. 
When an egg is fertilized, it develops into a larva which settles on some foreign 
object, where by repeated budding it grows into an asexual colony. This sequence 
is not followed by all members of the class, however, for some kinds, such as the 
Hydra of fresh water and the Sertularia so common along the Atlantic coast, produce 
special buds which are sexual. From them come young larvae without the interven- 
tion of a jellyfish stage. The graptolites are similar in appearance to Sertularia, and 
probably had a similar structure and much the same habits. 

Many layers of the black shales which form long narrow belts in the Great 
Valley of the Appalachians from eastern New York to Alabama show figures which 
suggest pencil markings. On closer examination it can be seen that these streaks have 
on one or both margins serrations which suggest the teeth of a saw. Linnaeus, a 
century and three-quarters ago, saw such markings on the black shale of southern 
Sweden and called them Graptolithus^ a word which has furnished a name for this 
group of fossils. Their leaf- and stemlike nature led the early paleontologists to 
class them as plants, but comparison with modern hydroids has revealed their true 
nature. Most of them are so flattened that it is practically impossible to learn what 
their real structure was. But, although they are largely confined to shale, it fortunately 
happens that some specimens preserved in an uncrushed condition have been found 
in limestone and chert, from which they have been freed by the use of chemicals. 
Gumbel, Holm, Wiman, Sollas, Bulman, and other European paleontologists have 
made elaborate studies which have furnished information that has explained the 

FIG. 15. A well-preserved graptolite (Isograptus) , twelve times natural size, 
with a portion enlarged to sixty-two times natural size. Note the details 
of the bilateral thecae, features which cannot be seen on flattened specimens. 
From O. M. R. Bulman. 

FIG. 1 6. A series showing the early growth stages of the modern hydroid, 
Endcndnnm. At left, the free-swimming ciliated planula; next, the cilia lost, 
the larva settles to the sea floor; there the basal disk is formed; later, enlarge- 
ment takes place, tentacles are formed, and (at right), the thin chitinous 
envelope bursts open. All from G. J. Allman. 

FIG. 17. Above, a pendent dendroid graptolite, Dictyonema, showing 
stages of growth, and an adult rhabdosome. Below, a many-branched axonoli- 
pan graptolite, with young specimens in the early stages of growth. From 
R. Ru edema nn. 

FIG. 1 8. Well-preserved specimens showing the early stages in tne growth 
of an axonophoran graptolite. At left, a sicula with virgella; n^xt, a sicula and 
first daughter theca; third, the stage in which a second theca has budded 
from the first; and at right, the stage in which the later thecae climb up the 
nema. From O. M. B. Bulman. 


common flattened specimens (Fig. 15). A considerable knowledge of the morphology 
and developmental history of the group has resulted. 

All graptolites form colonies which have the general appearance of recent hy- 
droids (Fig. 16). All seem to have been attached and to have had a basal disk or 
rootlike expansion for that purpose. From it springs a thin filament, or nema, which 
enlarges at the top to form the sicula, a thin-walled, chitinous cup containing the 
first zooid. From the first cup, or theca, arise one or more young individuals from 
which other buds grow out along certain definite patterns to form colonies of various 
shapes. The daughter zooids, like the parent, inhabited small bilaterally symmetrical 
receptacles which have a thickened ring about the aperture, one side of which in 
most cases is produced into a spine. The thecae, when flattened, form the sawtooth- 
like markings of the margins of the specimens in the shale. The bushy graptolites, 
the Dendroidea, have three kinds of thecae. There are large open ones, called hydro- 
thecae, which house the nourishing individuals; budding thecae, which do not open 
outward, and which give rise to the other kinds; and the so-called bithecae, of un- 
known function. Most of the graptolites, however, appear to have only one kind, 
the hydrothecae, which contained the feeding animals. 

As immature graptolites are common, the development of colonies, technically 
known as the astogeny, has been carefully studied in each of the three subdivisions, 
or orders, into which the group has been divided. The shrublike Dendroidea show a 
simple plantlike growth, the sicula first budding into a composite theca which consists 
of a nourishing and a budding individual (Fig. 17, above). The latter gives rise 
to other buds which produce elongate, straight, or irregular stipes, connected at in- 
tervals by crossbars which strengthen the colony. Most members of this order had an 
upright form of growth, and, to support the colony, secondary skeletal matter with 
rootlike expansions was secreted about the nema. Some, however, had a slender 
nema, and must have grown downward instead of upward. 

The simply branched forms included in a second order, the Axonolipa, grow 
from the sicula by the production of a bud which turns to one side, at right angles 
to the parent, whereas its daughter theca crosses to a position directly opposite to it 
(Fig. 17, below). From these thecae new individuals arise which may develop into 
a single pair of stipes, or may by further dichotomous budding produce four, eight, 
sixteen, or even as many as sixty-four branches, not connected by lateral supports. 
All species in this order have a long, slender nema entirely incapable of supporting 
the colony in an upright position. It is therefore inferred that such graptolites were 

Entirely different is the colonial development of the third order, the Axonophora. 
The sicula gives rise to a single theca, which turns abruptly aside and upward so that 
its aperture faces in the direction opposite to the original one (Fig. 18). Later buds 
keep the same orientation, producing a simple stipe with thecae on one or both sides 


of the nema. During growth the nema continuously elongates, becoming enclosed 
in the skeleton to form the axial support or virgula. Some representatives of this 
order are known to have had a bladderlike float; beneath this were clustered a number 
of reproductive sacs, in which larvae developed to the sicula stage. Some of the 
siculae remained attached to these gonangia, and the result was the production of a 
compound colony (synrhabdosome) with many simple colonies (rhabdosomes) pen- 
dent from one float. 

The habits and habitat of the graptolites have been much discussed, but what is 
now known of their structure, development, and distribution leaves little opportunity 
for difference of opinion. Such of the Dendroidea as have thick, rootlike supports 
appear to have lived in an upright position, growing like seaweeds on rocks, stones, 
shells, or other objects on the sea floor. All other kinds hung downward, many of 
them probably attached to seaweeds, whereas those with floats drifted about in the 
surface waters at the mercy of currents. Even those attached to seaweeds were in 
many cases a part of the migrant fauna, for great numbers of their hosts must have 
broken loose and drifted about. This interpretation of their mode of existence is 
borne out by their geographical distribution, for some of the species are almost cos- 
mopolitan, being found in such widely separated regions as Scandinavia, eastern 
America from Newfoundland to Alabama, Utah, Idaho, British Columbia, the Yukon, 
and Australia. It is to their rapid dispersal by oceanic currents that their great value 
in correlation is due. 

Why, if most of the graptolites led a floating existence, are their remains so 
generally confined to black or other shales of the finest grain? If they drifted about 
in the surface waters of the seas, they might be expected to die or be killed every- 
where, and fall to the bottom where all sorts of sediments were being accumulated. 
It is true that specimens are occasionally found in limestone, silt-stone, and sand- 
stone, but they are not abundant except in shale. Various possible explanations have 
been advanced by students of paleoecology, the science descriptive of the habits, habi- 
tats, and associations of extinct animals. The writer accepts the opinion championed 
by the first great student of this group, Lapworth, that the black shales of the Ordo- 
vician and Silurian represent muds accumulated at a considerable distance from 
shore, and that the presence or absence of graptolites depends upon the nature of the 
bottom on which they fell. It is probable that these animals were of more or less 
universal distribution in the surficial waters of the oceans and epeiric (intracontinen- 
tal) seas to which they had access, and that on death or detachment from their supports 
they sank to the bottom everywhere. Since the nema by which they were supported 
was very slender, every storm must have taken toll of great numbers of victims. 
Those which fell on a relatively firm bottom where starfish, cephalopods, snails, 
and other carnivores were present were quickly devoured, for their small size and 
thin tests made them an easy prey even to small animals. Such enemies were, how- 


ever, few or nonexistent in the soft mud of the offshore region. This is attested not 
only by the theoretical conclusion that animals would flounder and smother on a 
bottom composed of slimy ooze but by the actual fact that the shales containing 
graptolites seldom contain any other fossils than those which, like the graptolites, 
were planktonic (floating) in habitat. 

^ Enough appears to be known of the group to allow a brief summary of its history 
to be made. Bushy, upright, bottom-living forms were the first to appear, a few 

FIG. 19. Diagram to illustrate some of the stages in the evolution of the 
graptolites. A, basal portion of an upward-growing dendroidean. B, a pen- 
dent dendroidean. C, a many-branched axonolipan. D, an axonolipan with 
few stipes and with thecae turned downward. E, same, with thecae, and 
therefore stipes, turned upward. F, an axonophoran, with thecae turned 
upward. G, last and most specialized stage, an axonophoran with simple 
rhabdosomes, but thecae secondarily turned downward. An elaboration of a 
series by H. F. Cleland. 

specimens having been found in the Upper Cambrian (Fig. 19 A). Larvae of these 
seem to have settled in various places, some of them becoming accidentally attached 
to seaweeds. Those which became fixed on the under surfaces of fronds and those 
whose hosts were fragments of seaweeds which became detached and floated away 
found themselves inverted, and so grew downward instead of upward, thus originat- 
ing the pendent types (Fig. 19 B). Throughout the early history of the group there 
was a continuous reduction in the number of stipes. This may have been the first 
response to inversion and a somewhat unfavorable environment (Fig. 19 C-F). The 
zooids appear to have tried to turn toward the light; that is, they were, in the language 
of the behaviorist, positively heliotropic. Among the early and mid-Ordovician Axo- 


nolipa there is a general tendency for the stipes to bend upward as each new bud 
turned more and more toward the light (Fig. 19 D, E). The upright position was 
regained by the scandent Axonophora, in which all the thecae except the sicula open 
upward (Fig. 19 F, G). This was the most successful group, for, appearing in the 
upper part of the Lower Ordovician, it persisted till the end of the Silurian, whereas 
most of the Axonolipa disappeared at the end of the Ordovician. Had it not been for 
the accidental attachment to floating objects, it is likely that the graptolites would 
never have become abundant, for those which lived on the bottom, although existing 
until Devonian times, apparently were kept in check by the animals that preyed 
upon them. Those which went to the surface escaped their enemies and throve, 
especially after many of them had evolved floats of their own. But their doom was 
sealed toward the close of the Silurian, by the increase in numbers of actively swim- 
ming animals. These could eat graptolites faster than new colonies could grow, and 
the unfortunate unprotected hydrozoans seem to have escaped one sort of enemy 
only to fall prey to another. 

Considerable space has been devoted to this group because it is one of the few 
entirely extinct assemblages about whose rise and fall it is possible to draw relatively 
reasonable inferences. It also illustrates the methods of the paleontologists in dealing 
with an extinct group and, in this case, with material which on the whole is rather 


Methinks I see thee gazing from the stone 

With those great eyes, and smiling as in scorn 

Of notions and of systems which have grown 

From relics of the time when thou wert born. 

From a recently discovered poem, "To a Trilobite," 
by Timothy A. Conrad, first professional American 

Ever since people began to take an interest in "natural curiosities," the trilobites 
have excited the interest of those who have seen them in the rocks or in collections. 
Before the time of scientific study they were known as "petrified butterflies" or "flat 
fish," and their symmetrical forms, their elegantly ornamented surfaces, and the com- 
parative rarity of really complete specimens have made them the favorites among all 
invertebrate fossils to the present day. Moreover, these are truly ancient animals, 
which have been extinct for millions of years, although they were the dominant 
group in the period of the oldest really fossiliferous rocks. 

The name trilobite, or three-lobed stone, refers to the fact that longitudinal 
furrows down the back divide the surface of the shell into three lobes (Fig. 20). A 
more important tripartite division, however, is one in a transverse direction, since 
there are two conspicuous shields, situated at the anterior and posterior ends, con- 
nected by a segmented median portion. The cephalon or head shield of most bears 
a pair of more or less elevated compound eyes. The segments of the median portion, 
known as the thorax, are movable upon one another, like those in the body of a 
lobster or crayfish, and the hinder shield, or pygidium, has transverse creases or 
furrows that indicate an incipient segmentation. 

Most trilobites are small, the average size being perhaps about two or three 
inches. Some are only five or six millimeters in length when fully grown, but there 
are several giants more than twenty inches long. There was no steady increase in 
size among them, as there was among the Mesozoic reptiles and Tertiary mammals. 
Some of the largest are found in Mid-Cambrian strata, still more in the Mid-Ordo- 
vician, including the largest of all, length twenty-seven inches, an ancient inhabitant 
of France and Portugal. The Silurian has produced no real giants, but a considerable 
number of them appeared during Lower and early Middle Devonian times. Later 
trilobites are all small. 

The trilobites show all the sorts of visual organs to be found among the Arthro- 


poda. Some were entirely sightless. Many paleontologists think that in some groups 
this condition was primitive, whereas in others it can be shown that the blinding 
is secondary, for all gradations of degeneration, from forms with good eyes to those 
without any, can be traced. Some have a pair of simple eyes, with one lens apiece, 
and possibly a median simple eye, or ocellus, like that of many modern crustaceans. 
Still others have a group of three minute simple lenses on each cheek. Throughout 
the great superfamily Phacopidacea the large conspicuous eyes are of a type which is 
commonly spoken of as compound, but which might better be called composite or 

FIG. 20. A Mid-Cambrian trilobitc, Paradoxides, to illustrate the parts of 
the dorsal shield. Stippled portion at anterior end, cephalon; at posterior end, 
pygidium. a and b, cranidium (coarsely stippled) ; c, free cheeks (finely stip- 
pled); a, glabella with glabellar furrows; b, fixed cheek; d, visual surface of 
eye; e y genal spine; /, first thoracic segment; g, last thoracic segment (thorax 
not sdppled). The lines at the boundaries between the finely and coarsely 
stippled areas on the cephalon are the facial sutures. Original drawing by 
Eugene Fischer. 

aggregate. The eye is really made up of several simple ones, each with its own 
cornea. This is the sort of eye possessed by most arachnids. Each little lens is per- 
fectly distinct, separated from its fellows by narrow ridges. The composite eye differs 
in this respect from the true compound one, in which there is a common cornea 
and each minute element is so small that one has to put a specimen under the micro- 
scope to see the lenses. The latter is the type possessed by most trilobites. Some of 
these eyes are small; others are relatively enormous, the whole of the cheeks being 
given up to them. A climax is found in Symphysops, in which the eyes of opposite 
sides of the head are continued around the front. 

All trilobites are longer than wide, most of them with a cephalon both longer 


and wider than the pygidium, although there are some which are isopygous, that is, 
have subequal terminal shields. The number of segments in the thorax varies in the 
group as a whole from two to more than forty-four, although the number of species 
with more than twenty is small. There is a distinct relationship between the length 
of the pygidium and the number of thoracic segments. If the pygidium is long, there 
ara but few free segments; if it is short, there are many. This is due to the fact 
that the thorax is formed by the freeing of segments from the anterior end of the 
pygidium. The growing point, as in all arthropods, was immediately in front of the 
anal opening, which was situated near the posterior end of the pygidium. As in all 
arthropods, increase in size was accomplished at the time of the periodic shedding of 
the shell (molting), and it was at this time that new segments were introduced into 
the posterior shield in front of the anal opening and others freed at its anterior margin. 

The process of molting of modern arthropods is a critical one, for the animals 
must slough not only the whole external shell but a part of the lining of the anterior 
and posterior portions of the alimentary canal. Even the surficial covering of the eyes 
comes off. Growth proceeds to the point at which the animal becomes too big for its 
coat; then the shell bursts open down the back or along a margin, and the inhab- 
itant crawls out of its old case. It is then in what we call the soft-shelled condition 
we watch the season for soft-shelled crabs but the thin chitinous skin soon hardens, 
becomes more or less impregnated with salts of calcium, and the hard-shelled state 
is quickly resumed. 

Practically all trilobites with compound or aggregate eyes had a special provision 
for shedding the shell. Around the top of each eye where the edge of the visual 
surface joins the so-called palpebral lobe, there is a narrow groove which can be 
traced forward until it unites with its fellow groove on the other side of the cephalon. 
Traced backward from each eye, the grooves extend outward to the lateral margins, 
or backward to the posterior one. These grooves are the facial sutures. At the time 
of molting, the middle piece of the head (the cranidium) parted from the lateral 
portions (free cheeks), and the animal crawled out through the ample opening thus 
presented (Fig. 22) . Most of the specimens commonly found are cranidia, free cheeks, 
individual thoracic segments, and pygidia, the various members into which the shell 
disintegrated during ecdysis. Trilobites which are secondarily blind have obvious 
facial sutures. Most of those which seem to be primitively blind have them also, 
but they do not show on the upper surface, being marginal or submarginal in position. 
The free cheeks of some have become firmly ankylosed to the cranidium. Such trilo- 
bites seem to have molted as the horseshoe crab does, by splitting the shell along the 
margin of the head. 

Only the dorsal surface was protected by a firm shell. On the under side there 
was but a single plate, the hypostoma, which was situated beneath the central part 
of the head. The remainder was covered by a thin membrane. To protect their 


delicate lower surfaces, most trilobites, in the days after the Cambrian, acquired 
the habit of enrolling themselves. Many died in that position, and these furnish some 
of the best-preserved specimens. 

As has already been stated, the group was well represented even in Lower 
Cambrian times and perhaps reached its culmination in differentiation of species and 
abundance of individuals in the Upper Cambrian. Trilobites remained abundant until 
the initial days of the Mid-Devonian and then declined rapidly both in variety and 
abundance throughout the remainder of the Paleozoic. They survived until the 
Permian, where they are represented by perhaps a half dozen species. There are 
probably many reasons for their extinction. Other scavengers, particularly fishes, 
made competition for food increasingly severe. In the soft-shelled condition they 
themselves were excellent food, practically unprotected, a ready prey to the carniv- 
orous animals which became more and more numerous after the end of the Cam- 
brian. They appear to have withstood the attacks of the cephalopods, but failed before 
the keener-witted fish, which became gradually more clever and active while their 
prey was growing more stupid and logy. 

Since they are extinct, there has been a great deal of speculation about their 
affinities to modern animals. They were early recognized as arthropods because of 
their segmented bodies and compound eyes, and as they are always found associated 
with the sorts of animals which nowadays live in the sea, it was deduced that they 
were marine. It is now possible to prove their true relationships and to gain some 
plausible ideas as to their habits, for within the last forty-five years a series of fortunate 
discoveries has revealed the limbs on the under side of the shell. Although trilobites 
grew in a multiplicity of shapes, all appear to have possessed the same simple type 
of limbs (Fig. 21). Each animal had numerous pairs arranged all along the body 
from end to end, with little differentiation among them, those under the head shield 
being but slightly different from those under the middle or posterior part of the 
body. Each leg was bifurcated near the body, the upper branch (exopodite) forming 
a thin, fimbriated. organ, perhaps an external gill, whereas the lower one (endopodite) 
was a six-jointed slender walking leg, much like that of a thousand-legged "worm" 
or a wood louse. In some forms the segments of the lower branch were more or less 
flattened in a way that suggests that they may have been used in swimming. In all 
cases the limbs were attached to a series of processes which extended downward 
from the under side of the shell, beneath the two longitudinal furrows which give 
name to the group. Besides these divided appendages, all trilobites appear to have 
had a pair of many-jointed feelers (antennules) attached to the anterior part of the 
lower surface of the head; some of them possessed another pair at the posterior end. 

The first thing which strikes one in looking over a collection of specimens is that 
the shells are all relatively broad and depressed. Nowhere in the group do we find 
narrow, compressed animals like the sandfleas or shrimps. The upper surface is 


gently convex, the lower one concave. Moreover, all the organs were confined to 
the central one of the three lobes, the part outside the longitudinal furrows serving 
largely as a protective covering for the appendages. The broad form, suggestive of 
an overturned boat, was obviously well adapted to give the animal buoyancy, so it 
may readily be supposed that the creatures found it easy to keep afloat and easy to 
sv$m. On the other hand, it is equally obvious that they would not have made 
rapid progress in swimming, for they must be compared to a flat-bottomed scow 

FIG. 21. The anatomy of trilobitcs. At left, Neolenus, a generalized Mid- 
Cambrian form; schematized transverse section of a thoracic segment above, 
restoration of the entire ventral surface below. The endopodites are pur- 
posely left off the right side, to show the exopodites more clearly. At right, 
above, diagrammatic longitudinal section through the axial lobe of a trilobite, 
showing (stippled) the stomach and alimentary canal, and the segmented 
heart above it; beneath, a restoration of the lower side of the specialized 
Cryptolithtis. From P. E. Raymond. 

rather than to a keeled yacht. They lacked entirely the ideal lines of the swiftly 
moving fishes. The appendages, likewise, were not adapted to producing rapid and 
powerful strokes; their structure indicates that the trilobite proceeded rather slowly 
and sedately, with a kind of dog paddle. If the depressed form suggests buoyancy, 
it is also indicative, as is known from the study of modern animals, of bottom-living 
habits. Probably all could crawl by the use of the segmented portions of their legs. 
There is, however, a suggestion that the broad pygidium may have been used 
as a sort of flipper, which by spasmodic strokes gave a quick backward impetus to 
the body that sent it off in darting flight when occasion made escape necessary. The 
lobster executes such movements with its broad telson, and the construction of the 


trilobite is comparable. Action of this kind requires ample musculature, of the sort 
every epicure who has dissected the lobster knows that animal to possess. It might 
seem that nothing would be known of the muscles of the trilobite, since these are 
perishable tissues, but although the muscles themselves are lacking it is possible 
to reconstruct them from the scars of their attachments on the inside of the tests. 
Many of the fossils are found in an enrolled condition, and if they could roll them- 
selves up, it is obvious that they must have had the power of straightening out again. 
Hence they must have had two sets of longitudinal muscles, one dorsal and one 
ventral, and, as has been said, the scars of such muscles have been found. The size can 
only be inferred from the amount of room available to house them. It may be that 
trilobites with a wide median lobe had strong muscles, and the power of darting 
backward (Fig. 24, at right). Since the animal had no organs with which to steer, it is 
probable that the movements were rather erratic, and the trilobite was likely as much 
surprised as its pursuer by the course it took. 

The swimming trilobites probably lived near the bottom, for they had legs adapted 
for crawling as well as swimming. There are some, however, which seem to have 
floated and swum about in the waters at or near the surface of the sea, as members 
of the plankton. If a net of fine mesh be towed near the surface of the ocean at the 
present day it will catch a variety of small creatures, among them many crustaceans. 
In spite of their great variety, the animals of the plankton agree in having thin, trans- 
parent, or translucent shells, many of them spinose. One modern crustacean in par- 
ticular has received attention because of its general similarity to certain trilobites. It 
is a broad, flattened, spinose animal with large eyes. During the day it swims about 
in relatively deep water, out of reach of strong sunlight, coming to the surface only 
at night. Some of the spinose trilobites probably had these same habits; a particularly 
good case might be made out for one known as Robergia, which has a thin shell and 
large eyes, and is always found in fine-grained black shales, associated with graptolites, 
themselves a part of the plankton. 

Although most trilobites could probably both swim and crawl, some seem to have 
lost the power of swimming and to have lived entirely on the bottom of the sea. 
These animals are readily recognized by their elongate, wormlike form, the small 
pygidium, and the great number of segments between the two shields (Fig. 23). 
Some of the most ancient species show this form best, indicating that the habits were 
diilerentiated early. Others were not merely crawlers, but went further and bur- 
rowed in the mud or sand. Burrowing seems to have led to blindness, as might be 
expected from a complete adaptation to such a mode of life. Cryptolithus is the best 
example, because the most fully known. It is a small animal, from half an inch to two 
inches long, with a proportionately large and rather prettily ornamented head, the 
three mounds on which suggested its more commonly known name of Trinucleus 
(Fig. 21, at right). Around the .margin of the cephalon is a broad brim which long 

FIG. 22. Restoration, to show the probable appearance of a trilobite in the 
process of molting. Free cheeks were cast off along the facial sutures, and 
the animal crawled out Original drawing by Charles J. Fish. 



,,".", ,UaM 

FIG. 23. An elongate olenellid trilobite crawling on an irregular sea floor. 
Original drawing by Charles J. Fish. 

FIG. 24. At left, a Ncolemts walking toward the observer, on the "toes" of its bowed endop- 
odites. At right, above, an Isotelus jumping backward, using its pygidium as the lobster does 
its telson; below, swimming forward, by up and down movements of its posterior shield. Original 
drawings by Charles J. Fish. 

FIG. 25. A trilobite, Ccraunis, shown in feeding position above a gastropod. 
Original drawing by Charles }. Fish. 


ago suggested that the creature may have buried itself, like the modern horseshoe 
crab, by pushing down the brim and digging out the sand with its legs. The append- 
ages, when discovered, confirmed this suggestion, for they are stout, bowed at mid- 
length, and armed at the ends with numerous spines. Most striking of all, the feeling 
organs, instead of reaching forward, as in all other genera, were turned backward 
beneath the shell, so that they were out of harm's way in burrowing. But not all 
burrowers were blind, for in several groups there are species which have their eyes 
at the top of long, immovable stalks. These forms probably buried themselves up to 
the eyes and lay hidden with only their periscopes visible, perhaps awaiting their 
prey or observing the movements of their enemies. 

It has occasionally happened that specimens of fossil fishes and some other ani- 
mals have been found which had within them undigested but petrified remains of 
their last meal, so that it has been possible to get direct information as to the kind of 
food taken. The feeding habits of the trilobites, however, must be inferred from the 
structure of the jaws and the alimentary canal. As for the former, the animal was 
well supplied, for the inner segments of each leg were adapted to serve as jaws 
(Fig. 21, at left). We are used to thinking of these useful organs as being within 
the mouth, but they are external in many of the lower animals, and in this case some 
are far from the mouth. The basal segments of opposite pairs of legs projected inward, 
so that they met on the median line, and each was armed with numerous strong, 
sharp spines. The jaws probably waved backward and forward in unison, so that 
food seized by any pair of them would be swept forward to the mouth. 

Just what these animals ate is unknown, but since their jaws were neither large 
nor powerful, although numerous, it is probable that they lived upon small animals 
or plants or upon such tender matter as decaying organisms. Modern Crustacea are 
great scavengers, and it seems that their distant ancestors had the same habits. One 
can imagine a boat-shaped figure, coasting along the sea bottom, feelers outstretched 
fore and aft, eyes watching to port and starboard. Something is encountered and the 
animal settles over it, thus, by one action, reserving that particular morsel for its own 
refection and bringing its numerous jaws into position for work. The great underlip 
at the front is lowered, spreading the mouth open and making a trap past which 
food is not likely to pass. The distal end of the pygidium or its spines sink down 
into the mud, and dinner begins. Back and forth rasp the spiny jaws, and a con- 
tinuous stream of food moves forward to fill the hungry stomach (Fig. 25) . 

Not all trilobites fed in this way. Cryptolithus> the burro wer, was a mud feeder. 
Every schoolboy is familiar with the earthworm, which eats its way through the 
ground, taking out from the soil such organic matter as it can digest and voiding the 
remainder. Such were the habits of Cryptolithus, and probably many another of its 
kin. Many years ago, Count Sternberg, Barrande, and others who delved among the 
wonderful stores of fossils in Bohemia described the casts made by the mud which 


filled the digestive tube at death. By these and other fortunate circumstances we have 
come to know something of the internal anatomy (Fig. 21, at right, above). Such 
specimens show that the alimentary canal was exceedingly simple. A short throat 
passed upward to the saclike stomach,, which occupied the whole of the middle of the 
head. The stomach tapered into a narrow intestine extending back along the median 
line to the anal opening, on the lower side of the posterior end of the pygidium. 
Interesting deductions can be made from the circumstance that the median lobe 
of the head afforded lodgment for the stomach. That region, known technically 
as the glabella, is short and narrow in the more ancient trilobites; hence it may be 
inferred that they lived on concentrated animal food. In later ones this lobe increased 
in size till in some of the Devonfan forms it made up most of the head. Since bulky 
vegetable food requires more feeding and larger digestive chambers, it is supposed 
that the trilobites with large glabellae and, therefore, large stomachs were vegetarians. 
The middle of the head is, however, a most unfortunate location for the stomach, for 
the larger that organ, the less room for a brain, so we must ascribe to these unfor- 
tunates an increasing stupidity. In fact, one is tempted to impute a part of their 
downfall to the fact that there was no chance at all for cerebral development. For 
those who wish to point a moral, the case is obvious. 

Little more is known of the internal anatomy of trilobites than what has been 
set forth. On either side of the head there are glands which connect by lateral ducts 
with the stomach, a fact which suggests that their function was the secretion of 
digestive juices. A unique Russian specimen preserves the heart. It is a long, seg- 
mented, tubular body, like the heart of a worm or of one of the modern Crustacea. 
Its most remarkable feature is that it should have been preserved at all. An early 
attack of arteriosclerosis is almost the only explanation. 

Trilobites, then, seem to have been active, if stupid and inoffensive animals. 
They floated, swam, crawled, and burrowed, ate mud, carrion, seaweeds, and micro- 
scopic animals and plants. They played their part in the economy of the world in 
which they livecj and performed their humble service as scavengers countless years 
ago. Perhaps their greatest contribution is the aesthetic pleasure the contemplation 
of their elegant shells has given to countless collectors and students of fossils. 


And along came a spider, 
And sat down beside her . 

You know the rest of it. This is merely a subtle way of reminding you that a 
spider is an arachnid. But the spiders are not the only arachnids. "Daddy longlegs" 
is another, and so is the miserable "chigger," which the paleontologist collects in- 
voluntarily as he crawls over the deeply weathered Paleozoic outcrops of the South, 
hunting for fossils. All these are short-bodied animals, unlike the ancient trilobites. 
Entirely different in general appearance is the elongate scorpion, splendidly equipped 
posteriorly with a poisonous stinger. Although a terrestrial animal, it retains most 
of the features of its ancestors, which were entirely aquatic. 

The only primitively aquatic arachnids living today are the horseshoe crabs of 
the shores of eastern North America and southern and eastern Asia. They are com- 
monly known by the generic name of Limulus, correctly Xiphosura, although there 
are two other genera of xiphosurans. The horseshoe crab resembles a crustacean, and 
was long supposed to be one. It has a chitinous shell, compound eyes, and is aquatic, 
whereas the other modern arachnids are terrestrial and have aggregate eyes. The 
first appendages are, however, claws, not tactile organs. This is the simple way of 
distinguishing crustaceans from arachnids: first pair of appendages, tactile, Crustacea; 
first pair, claws, Arachnida. There are, of course, other obvious differences, such as 
that nearly all crustaceans have some biramous appendages, whereas those of arach- 
nids are uniramous. 

Limulus is conspicuously trilobate longitudinally, especially in its larval stages. 
It has several of the characteristics of trilobites, so many, in fact, that most zoolo- 
gists think that it descended from them. The chief difference in addition to those 
already cited, is that all the thoracic segments are ankylosed and form a rigid 
median shield. The appendages on the head are segmented, ending in claws or blade- 
like plates adapted for digging; these could have been derived from the inner 
of the two branches of the trilobitan limb (Fig. 26). Beneath the median shield, 
the legs are broad and flat. Dr. Leif Stoermer has recently shown that they have many 
of the structures of the trilobitan exopodite. In Limulus they serve as gills, which may 
have been their function in the trilobites also. 

The oldest definitely identified arachnids are found in Upper Cambrian strata, 
chiefly in Wisconsin. The best-known genus is Aglaspis, a form which much re- 


sembles Limulus, except that the segments between the head and the telson are free 
from one another. Dr. G. O. Raasch has found specimens showing appendages and 
has proved that the first pair are claws. The Middle Cambrian at the Walcott quarry 
contains animals of the same general shape and segmentation as Aglaspis, but these 
have tactile antennules and some of the appendages are biramous. They must, there- 
fore, be classified as crustaceans, but it is possible that they were ancestral to the 
Upper Cambrian arachnids, which, in turn, were distant ancestors of the horseshoe 
crab. Unfortunately,- no members of this line are known from Ordovician strata. 
The Upper Silurian of Scotland has furnished one well-preserved specimen much 
like a small horseshoe crab with free thoracic segments, and others are known from 
the Devonian. The group suddenly became rather abundant during Upper Carbon- 
iferous times, particularly in Great Britain and Illinois. Scattered specimens have 
been found in various other places, some of them in Permian beds (Fig. 26, at right). 
In fact, during the late Paleozoic the xiphosurans reached their culmination in 
variety, though not in size. All are small, an inch or two in length. They show an 
advance upon the condition of the Silurian specimen in that the body segments of 
most are coalesced to form a median shield, although some show obvious segmenta- 
tion. It is curious that all of the late Paleozoic specimens have been found in fresh- 
water deposits. The Upper Cambrian Aglaspis was marine; the Upper Silurian 
Neolimulus was found in a formation about which there is question. It contains 
some marine fossils, but also many others whose habitat is in dispute, as we shall 
see when we come to discuss the eurypterids. 

The subsequent story of the horseshoe crabs is quickly told. A few have been 
found in the marine Triassic; there is nothing more till the Upper Jurassic, when 
the modern Limulus appeared. Numerous beautifully preserved specimens have 
been found at Solenhofen. Some of them are larger than the modern Asiatic species 
but only half as large as the form so common along our New England shores. 

This is a strange history, with its transition from marine to fresh-water life, 
followed by a return to the sea. It seems probable, however, that some groups of fish 
may have had a similar experience. 

The xiphosurans may have been the first arachnids, but they made little impres- 
sion on early Paleozoic life. Their first cousins, the eurypterids, had the honor of being 
the only important arachnids during Ordovician, Silurian, and Devonian times. They 
were not shy or unobtrusive organisms, for some of the largest were from five to 
nine feet in length. On the average, however, they were much smaller, ten to twelve 
inches being the length of a good-sized specimen. 

Like the horseshoe crab, the eurypterid had an anterior shield with a pair of 
compound eyes and, near the median line, a pair of simple eyes or ocelli (Fig, 29). 
The appendages beneath this shield are also like those of Limulus, there being six 
pairs, the first of which are pincers by which food could be passed to the mouth 

FIG. 26. Ventral surface of a* modern Ldinulus, showing the six pairs of 
appendages on the anterior shield and the flattened ones on the mid-shield. I, 
anterior pincers (chelicerae) ; VI, the "pusher" legs, with their rosettes of 
flattened blades; O, operculum. From K. E. Caster. At right, ventral surface 
of the Permian Paleolimulus, with appendages. The chelicerae are not pre- 
served, but note the blades on the sixth legs. Three and a half times natural 
size. From C. O. Dunbar. 

FIG. 27. Stylonurus, showing the simple unspecialized appendages. From 
Clarke and Ruedemann. 

FIG. 28. The ventral surface of Euryptcrus. Note the chitinous ventral 
plates, as contrasted with the thin membrane of the trilobite. The anterior 
pincers are turned backward to the mouth. The next three pairs are diggers, 
and the last pair oarlike paddles. The large basal segments of this pair are 
the jaws. Between them is the operculum; back of it is the supposed ovipositor. 
From Clarke and Ruedemann. 


(Fig. 28). As in the trilobites and Limulus, the mesially directed projections of the 
appendages served as jaws. The limbs are variously modified in the various genera, 
four pairs behind the pincers being adapted for crawling or digging. Except in a few 
genera, the posterior pair, the sixth in the series, were the longest and largest, broad, 
flat, oarlike; useful in swimming. The trunk contains twelve free segments, without 
visible appendages. At the posterior end is a Limulus-like telson, a long spike in 
most, but broad and flattened in a few. The anterior part of the body consists of seg- 
ments which are distinctly wider than those on the posterior portion. The first two 
on the upper side are covered by a single plate below, a fact which shows that they 
are not continuous chitinous rings about the body. The skeleton actually consists 
of a series of dorsal plates (tergites) and ventral ones (sternites), the anterior pairs 
of which are not united at the lateral margins. When the ventral plates of some 
unusually well-preserved specimens were removed, remains of gills were found above 

Eurypterids do not show so many superficial resemblances to trilobites as do the 
horseshoe crabs, but they have an obscure trilobation, compound eyes, and free body 
segments. Unlike the trilobites, they have a chitinous ventral covering, most of 
which has already been described. On the head, behind the last and principal jaws, 
is an oval plate known as the operculum; behind it, on the median line, is an elongate 
organ whose function is really unknown, but which is supposed to be an ovipositor. 
This organ is pointed at the posterior end on some specimens, but on others that seem 
to be the same species, it is bifurcated. Since the latter individuals appear in general 
to be somewhat wider than the former, it is supposed that they were females. It is 
not uncommon, among modern arthropods, that the females are larger and broader 
than the males, but this is by no means so universally true that it can be accepted as a 
ready way of determining sex. Whether or not proportions have any significance, it 
seems probable that the organs on the median line do indicate differences in sex. 
Which is which is immaterial. The important point is that as early as Mid-Ordo- 
vician times, sex differentiation had achieved recognizable physical characteristics. 
These are, however, primary, not secondary features. The latter appear much later. 
Just when, nobody knows. 

The structure of the test (shell) of the eurypterid affords considerable evidence 
as to its habits. The flattened body indicates that it was a bottom-living (benthonic) 
animal. The pincerlike claws at the front show that it was a predaceous carnivore. 
All eurypterids were able to crawl, as is shown by the structure of the second to fifth 
pairs of appendages. Most could swim, as is indicated by the flattened shape of the 
sixth pair. Eurypterus itself, and some of its relatives, may have had the habits of 
Limulus. It could crawl and swim and, on occasion, burrow beneath the surface 
by the aid of its appendages and spikelike telson. But Pterygotus, the "Seraphim" 
of the Scottish quarrymen, was purely a swimmer and crawler, as is indicated by 


the marginal eyes and broad telson (Fig. 29). It could hardly have burrowed without 
injury to the cornea. Other forms, such as Eusarcus (Fig. 30), also had marginal eyes. 
Probably they were principally crawlers, for the short curved telson would have been 
no assistance in swimming. The interpretation of the habits of most of the various 
sorts of eurypterids is easy, but there is a line which culminated in the five-foot-long 
Devonian Stylonurus of the Catskills with appendages which seem useless (Fig. 27). 
All except the first pair are long and slender, each segment being unusually long. 
There are no oarlike paddles, and the telson is a spike. These facts make it evident 
the animals were not swimmers; on the other hand, the legs were too long and clumsy 
to have been useful in crawling. Nevertheless the group had a long and successful 

There is considerable difference of opinion as to whether the Ordovician and 
Silurian eurypterids lived in salt or fresh water. The conventional view, based on 
geological occurrence, is that they were originally marine but that in the late Silurian 
they became adapted to life in fresh water, where they persisted until their extinction 
in Permian times. We do not know their pre-Ordovician history, but infer that they 
sprang from some Cambrian arachnidian stock closely allied to Aglaspis. One of the 
most conspicuous of the Mid-Cambrian fossils from Burgess Pass is an eurypterid-like 
creature which Walcott named Sidneyia after his young son, who found the first 
specimens. Although the shell is eurypterid-like, the animal had tactile antennules, 
and the appendages are divided like those of a trilobite. It is chiefly remarkable for 
a pair of extraordinary claws. This creature, however, was a crustacean, and so 
specialized that it could not have been ancestral to the eurypterids. Its presence in 
Mid-Cambrian rocks proves that there was at that time a line of crustaceans in which 
evolution was tending toward the elongate body form of the eurypterids. There is 
even a differentiation in the trunk between anterior broad and posterior narrow 
segments. This sort of animal suggests, although it by no means proves, a marine 
origin for the eurypterids. Did the ancestors migrate into fresh waters as early as 
Cambrian times? 

Dr. Marjorie O'Connell, while studying under Professor A. W. Grabau, wrote 
a long paper in which all the then known occurrences of eurypterid remains were 
discussed. The evidence cannot be reviewed in detail here, but her conclusion was 
that all were inhabitants of fresh water. She pointed out that no complete specimen 
had ever been found in the Ordovician, which might indicate that the animals lived 
in rivers and that only their broken and macerated fragments were carried to the 
sea, to be recovered by paleontologists from marine deposits. 

The climax of eurypterid differentiation was in the Upper Silurian. Rocks of 
this age have produced the best and largest specimens, and in them are the four 
famous localities for such animals. At all of them the conditions that existed while 
the strata were being deposited were probably somewhat unusual, for, although 

FIG. 29. Pterygotus, the largest and most aggressive of the eurypterids, and 
probably the largest arthropod which ever existed. Note the huge size of the 
pincers, the feeble development of the diggers, the small size of the paddles, 
and the large flat telson. From Clarke and Ruedemann. 

FIG. 30. At left, a Silurian Eusarcus. At right, sketch of a modern scorpion 
to show similarity of trunk and curved telson. Eusarcus from Clarke and 
Ruedemann, scorpion after R. I. Pocock. 


mollusks and a few other marine animals are tound in the same layers, the faunas 
are admittedly peculiar. The writer has been fortunate enough to have visited all 
the sites, and finds that the eurypterid-bearing rocks were not all formed under the 
same conditions. The first site is a tiny, now disused quarry a half mile southwest of 
the center of the hamlet of Kichelkonna on the western margin of the island of Oesel 
(now Saaremaa) in Estonia. Here the specimens are wonderfully preserved in a 
light gray dolomite. So remarkable is the preservation that some individuals retain 
their color markings; practically all have the original chitinous shell. It is thinner 
than the thinnest tissue, and is the despair of the collector, for as the newly excavated 
specimens dry, it breaks, curls up, and blows away. Gerhard Holm, an eminent 
Swedish paleontologist, discovered how to remove the shells from specimens found 
at this locality, and how to preserve them. To him is due a great deal of our knowl- 
edge of the morphology of the eurypterids. Curiously enough, he never disclosed 
the secret of his technical skill. Typical marine beds are found both below and above 
this eurypterid bed. The lithology is the same in all. Another famous locality is 
south of Lcsmahagow, not far southeast of Glasgow, in the district where the banks 
of the Clyde are really "bonny." Here the rock is a dark gray shale, which contains 
so many marine mollusca that there can be no question but that it was deposited in salt 
water. It does grade, higher up, into r>stracoderm-bearing beds about which there 
may be some question. The third and fourth localities are in New York state, one 
in Herkimer County, south of Utica, and the other in the southern part of the city 
of Buffalo. Most of the good American specimens to be seen in museums came from 
one or the other, although neither is now an active producer. The rock at both is 
an argillaceous dolomite. Few other fossils accompany the eurypterids, although 
Dr. Rudolf Ruedemann has pointed out that such as have been found are marine. 

In Herkimer County and at Buffalo the Eurypterus-bearing beds are almost the 
youngest Silurian strata, but in the region between, centering around Syracuse, are 
the great Upper Silurian salt deposits and gypsiferous shales. These beds, perhaps, 
suggested the idea that all Upper Silurian deposits are abnormal. To account for 
the salt and gypsum it has been necessary to consider this part of New York to have 
been, at the time, a region of embayments, nearly cut off from the sea, in which 
evaporation produced concentrated brines from which salt was deposited. Into 
such embayments may have been swept such eurypterids as could not maintain 
their position in the tributary rivers. Death in the salty water would account for the 
abundance of well-preserved, complete specimens. 

It is impossible to refute certain arguments. If the remains are fragmentary, it is 
because the animals lived in rivers, and only remnants of molted specimens reached 
the sea. If the individuals are complete, it is because they were killed when they 
were washed out of their normal habitat into saturated brines. 

The discussion of these localities may serve to show that conditions were not 


the same in Oesel, Scotland, and New York. At all of them well-preserved specimens 
are found. If only the one at Lesmahagow were known, it might have been in- 
ferred that it was the increasing freshening of the waters which caused the destruction 
of the eurypterids. 

Whatever their trials and troubles, eurypterids survived them and went on to new 
triumphs in fresh waters during Devonian times. Perhaps they grew stronger through 
adversity, for their shells were no longer carried into the sea, even after death. We 
have no idea what caused their final extinction. Post-Devonian specimens are rare, 
and relatively small, even though the group survived till the Permian. Perhaps the 
early amphibians and reptiles ate them. 

Although the eurypterids have vanished, they have left us a souvenir, the scor- 
pions. If one examines a scorpion, preferably a dried or pickled one, he sees that it 
is practically a eurypterid (Fig. 30, at right). The anterior shield bears the same 
number of appendages, the first a pair of pincers, all the others walking legs. 
But the eyes are aggregate, not compound. The trunk has the same number of 
segments, divided, as Sir Ray Lankester, the most distinguished naturalist England 
has produced in recent years, liked to say, into tagmata. That is, the anterior 
ones are enough broader than the posterior to define separate regions. At the 
posterior end is a telson, a short segment with a curved spine, through which the 
amiable animal delivers its poison. Beneath the ventral plates of the anterior seg- 
ments of the trunk are the breathing organs. They are similar to those of the euryp- 
terids, but modified for the absorption of oxygen from the air, not from the water. 
They are, therefore, not gills but lungs, denominated, because of their leaflike struc- 
ture, book lungs. Air is admitted to them through paired circular openings known 
as stigmata. 

The oldest known scorpions are three individuals found associated with euryp- 
terids, two in Europe and one in New York. They are known to be scorpions and 
not eurypterids because they have pincers on the second pair of limbs and "combs" 
on the first trunk segment. The latter appendages are peculiar to scorpions. These 
specimens were for long celebrated as the oldest air breathers, but Dr. R. I. Pocock 
of London has shown that they lacked stigmata and hence were probably as fully 
aquatic as their associates. If one compares the skeleton of one of these Silurian scor- 
pions with that of the contemporary eurypterid, Eusarcus, he is readily convinced of 
the close relationship between the two groups (Fig. 30). It must be admitted that 
the connecting links have not as yet been discovered, but there can be no doubt where 
the ancestry of the scorpions lies. No Devonian members of the group have yet been 
found, but Carboniferous scorpions have stigmata. They are unquestionably air 
breathers. The successful branch of the family was the one which learned to do 
something new. 

Once the arachnids got out into the air, evolution was rapid. Spiderlike animals 


have been found in the Lower Carboniferous, and true spiders, as well as various 
related forms, are known from the Pennsylvanian. Nothing is yet known of the 
ancestry of these groups. Their secrets are still locked in undiscovered layers of the 
Devonian deposits. Specimens found in the Rhynie chert of Scotland show that mite- 
like creatures were in existence as early as Mid-Devonian times, but their aquatic 
ancestors have not yet been discovered. Almost all modern arachnids are air breathers. 
It is somewhat exasperating that so little is known of their origin. But it is not 
really surprising, for comparatively few terrestrial animals die in situations where 
they have a chance of being preserved as fossils. The more completely terrestrial the 
animal, the less chance the paleontologist has of finding its remains. Judged on this 
basis, one would suppose that the remains of the flying insects would be much rarer 
than those of the wingless arachnids. But such is not the case. Hundreds of specimens 
of fossil insects may be found in strata which yield only a single spider or spiderlike 
creature. The inference is that the early air-breathing arachnids lived in moist 
tropical and subtropical jungles where the decay of all organic material was rapid. 
Many of them probably wandered far from the streams and swamps which seem 
to have been the haunts of many ancient insects. Little is known of their ancestral 
food habits. The fondness of spiders for the juices of insects may have been initiated 
by feeding upon helpless larvae, found in the decaying logs of the jungles. Perhaps 
the question: 

"Will you walk into my parlor?" 
Said the spider to the fly, 

was not asked until comparatively recent times. Flies wrapped in webs and individuals 
partially devoured have been found associated with spiders in the Oligocene amber 
of the Baltic but in no older deposit. 


Their strength is to sit still. 

Isaiah, xxx, 7 

Although we no longer class together all animals with radial symmetry, it is 
convenient, for the purposes of this book, to treat of them in one chapter. Coelen- 
tecates and echinoderms have it in common that in general they thrive best in warm 
water, and that both have contributed greatly to the formation of limestone. In 
speaking in these general terms I am, of course, referring chiefly to the attached 
coelenterates, the corals, and to the sessile echinoderms, the cystids, the crinoids, and 
the blastoids. 

The animal or polyp of a coral is simple, but not quite so simple as the hydroid 
mentioned in the chapter on graptolites. It is, in fact, nothing but a living tube, with a 
mouth surrounded by tentacles, but there is a short oesophagus leading into the body 
cavity. The latter is divided by incomplete radial partitions (mesenteries) into a 
series of alcoves which afford a certain amount of privacy to the digestive processes 
(Fig. 12 D). These infoldings of the walls greatly increase the amount of endoderm, 
whose cells attend not only to the processes of digestion and assimilation but also 
to reproduction. As in the sponges, many of the cells have cilia, for each is practically 
a protozoan. The animal is not, however, a mere colony of protozoans, for there is a 
division of labor. The components are not all alike. Some are sensory, some digestive, 
some reproductive. Although the polyps have nothing remotely resembling a brain, 
or even a central nervous system, they get along very well. All are carnivorous. 
Perhaps you have seen pictures of a sea anemone swallowing a fish, an animal a 
thousand degrees above it in the social scale. It is not necessary even for the anemone 
to pursue its prey; its strength is to sit still. 

But sitting still in warm sea water saturated with calcium bicarbonate has its 
disadvantages. The animal has to get rid of the surplus calcium. The coral does this 
by secreting, or excreting, a skeleton beneath, rather than around its body. The larva, 
after a short free-living existence during which it passes through a part of its de- 
velopment, settles on some hard object upon which the remainder of its life is to be 
spent. The lower surface then puckers up into a series of radial folds, alternating in 
position with the mesenteries. In the grooves so formed, minute irregular bodies 
(spicules) of calcium carbonate are secreted. These eventually coalesce, producing 
a set of radially arranged plates (septa). With further growth, new folds appear at 


the base, and new septa are formed between the primary ones. As the process con- 
tinues, more spicules are added at the outer ends of the septa, along the outer surface 
of the tubular polyp. Thus the ends of adjacent septa become connected, and a 
circular outer wall (theca) is produced. Later in life most corals continue to form 
new septa and new thecal tissue, so that a conical skeleton is formed. It should be 
noted, however, that this is not a house into which the animal can withdraw. The 
coral never gets beyond the stage of building the foundation. To make sure that 
there shall be no retreat, the animal builds more or less horizontally arranged plates 
extending from wall to wall (tabulae) or from septum to septum (dissepiments). The 
living coral is always on the uppermost story of his pedestal; one can scarcely call it 

Some corals reproduce sexually only, producing simple cup corals, which were 
much more numerous during the Paleozoic than they are today. At the present time 
these are found most commonly in cold water, either deep-sea or, if shallow, in sub- 
arctic or cold-temperate regions. As has already been pointed out, reproduction by 
budding was unusual in Cambrian times. It is shown there by only a few animals, 
sponges in the Mid-Cambrian, graptolites in the Upper. But by the Mid-Ordovician 
it became a commonplace, probably the result of many generations of fixation. Per- 
haps the warmer waters and more abundant food had something to do with it. 

Most modern corals and a large proportion of the Paleozoic ones are colonial in 
habit. As new buds form, they build their skeletons upward and outward, producing 
more or less hemispherical, moundlike masses. Much more rare in the Paleozoic are 
colonies of the "staghorn" type so common at the present day. The individual cups 
(corallites) in the colony (corallum) may be circular if the buds are sufficiently diver- 
gent. Most, because of rapid budding and consequent crowding, are polygonal in 
section, three, four, five, six, or seven sided. Some achieve the optimum for closely 
arranged polygons, the hexagonal section. The square and the hexagon are the only 
patterns shown by colonies in which all the corallites have the same shape, but these 
are less common than corallites which have a variable number of sides. 

The stony corals are divided into two great groups, the septate, with conspicuous 
radial septa, and the aseptate, without them. To the first belong the common reef- 
building corals; to the second, the sea fans, dead-men's-fingers, and the precious coral. 
Both were represented in the Paleozoic, although the aseptate forms then living 
differed from those of today. Although the septate kinds of the Paleozoic and 
Mesozoic seem more alike, they are really structurally different. The Paleozoic ones 
have four primary septa, to which others are added in multiples of four; hence they 
are called Tetracoralla. In modern corals there are twelve (specialists please read this 
number as two, four, six, or twelve) primary septa, and the increase is in multiples of 
six. Naturally they have been named the Hexacoralla. This difference in numbers 
may not seem to be a matter of fundamental importance, but it really is, as may be 


seen if the mode of introduction of new septa be studied. In the tetracorals they origi- 
nate as outgrowths from two sides of one of the primary septa (the cardinal) and from 
one side of each of two lateral ones (the alar). It may be seen by consulting the 
accompanying figures that, although the symmetry appears to be radial, as the corallite 
is viewed from above, it is fundamentally bilateral, as is shown by the structure of the 
cardinal side (Fig. 31). Young hexacorals likewise have bilateral symmetry, but it is 
lost so soon after the introduction of the first septa that the radial arrangement seems 
to be primary. It has already been pointed out that the graptolites differ from modern 
hydroids in being bilaterally symmetrical. The same is true of the early septate corals: 
in fact, they remained bilateral till the end of the Paleozoic. The evidence indicates 
that radial symmetry is secondary, the result of fixation. Spherical symmetry may be 
more primitive than bilateral, but the days of spherical symmetry in adults appear to 
have been over before the beginning of the Cambrian. 

FIG. 31. Diagrams of a tetracoral showing, A, the divergence of septa 
from the primary one (c) on the cardinal side, and, B, from the alar (a). 

The aseptate Paleozoic corals differ from most modern members of the group in 
that their skeletons were external, not internal. No sea fan, gorgonian, or bit of 
precious coral has been found in Paleozoic rocks. In their place are Tabulata, such as 
the honeycomb-like Favosites (Fig. 38), the chain coral, Holy sites, and others familiar 
to everyone who has ever seen a textbook of geology. Attempts to explain their rela- 
tionships to modern corals have been and still are being made. We need not go into 
the matter. Perhaps they were the "lower classes" in the corallian social scale. Each 
individual produced a multitude of offspring who remained at home, forming large 
colonies. With so many mouths to feed, no individual got enough nourishment to 
grow large, for the number of buds was enormous. Some coralla of Favosites are five 
or six feet in diameter. They compare well with the "brain corals" of the present day, 
forms in which the incomplete separation of the daughter polyps from the parent 
produces convolutions superficially similar to those of the human brain. 

It is interesting to note that the corals, which had a modest beginning in the 
Ordovician, became sufficiently abundant in Silurian and Devonian times to build 
structures analogous to modern coral reefs, though no cases are known in which such 
reefs corresponded exactly to modern barriers or atolls. It may be that this adaptation 
caused the extinction of tetracorals and Tabulata. It is known that the reef-building 


corals of the present day thrive only in waters whose mean annual temperature does not 
depart greatly from 68 F. A drop of a few degrees will kill them, if maintained for 
any considerable period. The reef corals also succeed best in relatively shallow water, 
that is, from just below low-tide level down to a depth of about twenty to twenty-five 
fathoms. Not all hexacorals are of this type, and not all of them require such a relatively 
high temperature. Some live in the deep sea, and some in high latitudes. Corals are 
found near Woods Hole, Massachusetts, and off the Newfoundland banks, and at least 
one species grows abundantly off the coast of Norway in rather deep water. These, 
however, are not reef-builders. 

Professor R. A. Daly has set forth considerable evidence to show that during the 
Pleistocene glacial period, most of the reef-building corals of the Pacific islands were 
killed and that the luxuriant growth on the present reefs is due to a later re-coloniza- 
tion. The great masses of ice which covered the polar regions, extending southward 
to a latitude of 40 on the continents, must have had a chilling effect upon the oceans. 
That it was a time of general refrigeration is indicated by the presence of local glaciers 
even in the Hawaiian Islands. Since a small change in temperature affects the 
corals so seriously, life on many of the reefs may have been extinguished during the 

In the light of this it seems probable that the almost total change in coral fauna 
at the end of the Paleozoic was due to the Permian glaciation. At that time unusually 
cold conditions existed both north and south of the equator, as is indicated by glacial 
deposits in South Africa, South America, India, and in the small area near Boston 
made known by Robert W. Sayles. The accumulation of ice and the general lowering 
of temperature must have had their effect upon the oceanic waters. The tetracorals 
and Paleozoic aseptate corals may have been exterminated. These animals doubtless 
throve best in warm localities, for their skeletons are most abundant in limestone, and 
it is well known that at the present time marine limestone is accumulating chiefly in 
warm water. 

It may be supposed that skeletonless Hexacoralla were in existence during the 
Paleozoic but that they occupied the colder waters, holding the same position in the 
community that the subarctic corals do today. Since they were accustomed to a low 
temperature, they would easily survive the chilling effects of the Permian glaciation. 
In fact, this may have been the moment of their great opportunity. The destruction 
of the earlier warm-water corals left vast regions suitable for colonization. Current- 
carried larvae no longer perished because of transportation into warm seas. Their 
ecological position was vacant. Food was still abundant. The time which comes once 
in the lifetime of every race, as well as in that of every man, had arrived. The naked 
actinians spread into every sea. 

Is there any indication that this happened? Possibly a clue, to which the writer 
called attention some years ago in an article which excited much interest but which 


has never been favorably reviewed. The oldest known coral is an actinian, that is, a 
skeletonless sea anemone, of the Mid-Cambrian. Walcott described it by the name of 
Macfenzia as a sea cucumber (holothurian), but Drs. A. H. and H. L. Clark soon 
showed that it was not an echinoderm (Fig. 32). Like other marvelously preserved 
specimens from Burgess Pass, its photographic imprint shows that it had internal 
mesenteries and the remains of sixteen retracted tentacles about the mouth. To all 
intents and purposes it is identical with Edwardsia, a sea anemone still common in the 
cool waters about the English coast. Despite its sixteen tentacles and mesenteries, 
Edwardsia is considered by students of corals to be a member of the Hexacoralla, 
because the more advanced members of that group pass through an Edwardsia stage 
in their individual development. Since the Mid-Cambrian Macfenzia and the modern 
Edwardsia have the same important characteristics, it seems probable that similar 
corals must have existed ever since Mid-Cambrian times. Their lack of skeleton seems 
to be due to the fact that they were inhabitants of cold water. But suppose that 
during the cold-water period of the Permian, the Mac\enzia-Edwardsia type of animal 
spread allpver the earth. As the waters warmed up again during the Mid- and Late 
Triassic, they would have been forced to secrete calcium carbonate. Might not the 
skeletons have been of the hexacorallan type, since even now the Hexacoralla pass 
through an Edwardsia-Maclfenzia stage? 

FIG. 32. A, diagram of the modern actinian, Edwardsia. B, larva of a 
modern actinian in the eight-mesentery stage. C, sketch of Macfanzia, with 
its sixteen contracted tentacles. From P. E. Raymond. 

The history of the aseptate Paleozoic corals is even more obscure. The most 
successful type, the favositid, culminated and perished in Mid-Devonian times. Rela- 
tives struggled on till the late Carboniferous. Perhaps the cold waters of the Permian 
finally "did them in." One can see that the muddy waters of some of the shallow late 
Devonian seas may have exterminated Favosites, but it is remarkable that the larvae 
of such a virile stock should not have managed to get into the clearer waters. There 
may have been "barriers" of which we as yet know nothing. 

Except for their radial symmetry, echinoderms have no resemblance at all to corals. 
Although many are sessile, they do not reproduce by budding. There is an alimentary 


tract hung within the body cavity, as in all coelomates, and there is a central nervous 
system. The skeleton, which consists of calcareous plates, is internal rather than 
external, though close to the surface. It is deposited by special cells which build up 
cribriform plates composed of spicules of calcium carbonate. Although the plates of 
modern echinoderms are porous when deprived of their organic matter, each plate or 
spine has the structure, though not the form, of a single crystal; when broken, a typical 
calcite cleavage is shown. This applies to the fossils as well. So far as is known, 
no other sorts of organisms have this property. Often it is possible to identify 
the fragments of a badly preserved specimen as an echinoderm by its rhombohedral 

The starfish is the most familiar member of the group. Although free through- 
out adult life, it is sessile for a short interval during the larval stages. A brief discussion 
of its structure will assist the interpretation of the ancient forms. It shows conspicu- 
ously one of the most characteristic of echinodermal structures, the water-vascular 
system. On the upper surface, between two of the rays, is a perforated plate, the sieve 
plate or madreporite, through which water is drawn into a tube which leads to a 
circum-oral vessel (the stone canal) on the lower side. Extending along a median 
groove on each ray, externally, is a branch from the stone canal carrying water to the 
"tube feet," which are the prehensile and locomotory organs. Each tube foot is con- 
nected through a pore between adjacent pairs of rafterlike plates (ossicles) with a 
small spherical reservoir within the arm. The ossicles, like the tube feet, are paired 
on either side of the median groove, forming a long, narrow, outwardly tapering 
region known as an ambulacral area. At the tip of each ray, at the end of the median 
groove, is a small plate bearing a pigmented organ, probably sensitive to light and 
hence called an eye. From it the plate through which it projects has been named the 

Naturally, all directions from the center of a radially symmetrical animal are 
radial, but technically the ambulacral areas of the echinoderm are considered radial, 
whereas the regions between them are interradial. Thus the madreporite of the star- 
fish is interradial, the oculars are radial in position. The terms "dorsal" and "ventral" 
have little or no meaning when applied to the echinoderms. The mouth is beneath 
(ventral) and the anal opening on the upper side (dorsal) in the starfish, but in many 
others these organs are on the same side, dorsal in some, ventral in others. It is, 
therefore, customary to speak of the side having the mouth as actinal; whatever is 
opposite to it is abactinal. 

The starfish living at the present day are of two general kinds: some, such as the 
common ones of the New England waters, have a flexible skeleton consisting of 
small plates embedded in a tough integument; others, inhabitants of warmer regions, 
have larger plates which greatly reduce the mobility of the arms. Both kinds have 
existed since the Mid-Ordovician, the time when the group made a sudden appearance. 


An early narrow-rayed line of starfish culminated in Mississippian times in the true 
brittle stars, ophiurans, in which each pair of plates from opposite sides of the median 
groove has been combined into one so-called "vertebral" ossicle. 

Starfish do not have branching arms, but some modern ophiurans, the basket 
stars, do. A basket star, turned actinal side up, bears considerable resemblance to one 
of the feather stars, free-swimming crinoids (Fig. 33, at left) . The latter animals are 
unknown on American shores. Probably more Americans are familiar with fossil 
than with recent crinoids, for quarries at Ottawa, Ontario; Warsaw, Indiana; Burling- 
ton and Keokuk, Iowa; and localities in New York, Ohio, Tennessee, Missouri, and 
elsewhere have filled the collections of museums and private individuals with beauti- 
ful specimens. In fact, fossil crinoids were known to naturalists before recent ones 

FIG. 33. Diagrammatic representations of, at left, a modern free-swimming 
crinoid; at right, a modern stalked crinoid. From A. H. Clark. 

were discovered. All the localities mentioned above furnish only the old-fashioned 
stalked crinoids. Feather stars are of relatively modern invention, dating, so far as is 
now known, from the Jurassic. Such free-swimming forms are now most common 
in shallow tropical waters, though one hardy form is found in the Irish Sea. Their 
stalked cousins, so abundant in Paleozoic shallow seas, are now few in number and 
confined to restricted areas, although not to deep water, as is generally supposed. 

Modern crinoids, whether free or fixed, have the same structure. The mouth 
is, on the upper side, in the midst of the appendages. From it radiate five primary 
grooves, with branches extending onto the arms; they are lined with cilia which 
capture food and carry it to the mouth. Branches of the arms may be few or numerous, 
but all are made up of calcareous plates, connected at the ends, deeply grooved above. 
Along their sides are upright outgrowths, the pinnules, which are not food-getters but 
which bear the sexual products (Fig. 33). The main portion of the body, chiefly a 
spirally twisted digestive system, is enclosed in a calyx consisting of two or three 
horizontal rows of calcareous plates. Those from which the five primary arms origi- 


natc are known as the radials (Fig. 34, at right). Below, and alternating in position 
with them, hence interradial, are the basals. Below the basals of some crinoids there 
is a third row of five plates, radial in position, the infrabasals. The infrabasals, if 
present, or ~basals otherwise, rest upon the dorsocentral abactinal plate of free crinoids 
or upon the upper columnal of the stalked ones. The column itself is made up of a 
series of coin-shaped plates with a central opening or lumen. They are attached to one 
another by ligamental tissue situated in radial grooves, a method of union which gives 
flexibility to the stalk. Some crinoids, chiefly the modern ones, have lateral appendages, 
cirri, on the stalk or on the dorsocentral plate. They serve the free crinoids as anchors 
during temporary "tie-ups." 

Modern crinoids of the stalked variety have been in existence since Triassic times. 
All the Paleozoic ones (there was greater variety in those days) differ from them in 
that the mouth does not open at the surface but is completely covered by a series of 
plates forming a vault known as the tegmen. Like the later crinoids, the Paleozoic 
ones gathered food on the arms, but it passed along food grooves which led through 
lateral openings at the bases of the arms to the concealed mouth (Fig. 34) . Some of 
the Paleozoic crinoids have a simple calyx, composed of two or three circles of plates. 
But more of them have the primary calyx (patina) enlarged by the inclusion of 
proximal portions of the arms. Primarily, there are but five arms; if more issue from 
the sides of the calyx, then its upper portion has been secondarily enlarged by the in- 
corporation of arm plates (brachials) and interradial plates (interbrachials) (Fig. 34, 
at left). The most abundant Paleozoic crinoids are the Camerata, forms with large 
secondary calices including many brachial and interbrachial plates below the places 
of attachment for the free portions of the arms. They also have large, highly vaulted 
tegmens. Perhaps the most conspicuous feature is the elongated anal pyramid or tube. 
The anal opening of many is within the circle of arms, a strange arrangement for 
which their ancestors, the cystids, were responsible. Naturally this was no place for 
an excretory organ, at the very center of the food-collecting grooves. There wasn't, 
however, much that the unfortunate animals could do about it. But accidental varia- 
tion, or inheritance, or an orthogenetic urge, or something, attempted to improve 
conditions. The anal pyramid, primarily a circle of five or ten plates, rose higher and 
higher, till at last it towered far above the arms. From its lofty and imposing height 
it could now shower out the waste where all branches of the food grooves, rather than 
merely the inner portions of them, could collect it. Evolution is not in all cases in- 
telligently directed. Perhaps this practice of using the same food repeatedly may have 
been a factor in the sudden downfall of the crinoids. 

The crinoids, like the starfish, appeared first in the Middle Ordovician but reached 
their climax during the early part of the Mississippian, when they were not only 
abundant but highly varied. This, however, was not the only time when they were 
abundant, for strata of various ages consist more or less completely of disjointed 


columnals and plates. Since the skeleton is held together largely by ligamental tissues, 
the parts naturally became separated after death. One of the reasons why the early 
Mississippian strata contain so many well-preserved specimens seems to be that among 
the camerate crinoids there was a constantly increasing tendency for the adjoining 
plates of the calyx and tegmen to become cemented along their edges and thus more 
resistant to disintegration. 

Why so many kind of crinoids disappeared in Mid-Mississippian times is not 
known. Before the Mid-Pennsylvanian they had become rare, and they have remained 
in a subordinate position since, although after achieving freedom from the stalk in 
Jurassic times they "picked up" a bit. Even the free crinoids may have passed their 
maximum, for their Mid-Cretaceous representative, Uintacrinus, is by far the largest 
known. Individual calices are as much as three inches in diameter, and arms have 
been traced to lengths of forty inches. Kansan slabs completely covered with beauti- 
ful specimens of Uintacrinus with intricately intertwined arms tell of Mid-Cretaceous 
disasters; not, surely, due to panic in fire or flood; nevertheless, the result of some 
unreasoning crowding. 

The nut-shaped blastoids are the most stereotyped of echinoderms. All are alike 
in that they have a slender, but rarely preserved, column, a calyx composed of three 
or five basals, five deeply cleft radials (forked plates), and five small interradial 
"deltoids." Radiating from the mouth are five broad, ambulacral areas, which during 
the life of the animal bore many simple pinnules (Fig. 37). Surrounding the plate- 
covered mouth are paired openings connected with the water-vascular and genital sys- 
tem. An unpaired opening, or in some a specially modified member of the paired 
ones, is the anus. The water-vascular system, which left but slight marks upon the 
skeleton of the crinoid, was highly developed in the blastoids. 

Much has been, and much of general interest could be, written about the group, 
but there are limits, even to books. True blastoids appear first in the Silurian. The 
various blastoid-like animals of the Ordovician are so complicated in structure that 
they must for the present be left to the specialist. There are several Middle and Upper 
Devonian species, but it was not until after the decline of the crinoids in the early 
Mississippian that this group became common. They apparently took over the vacated 
ecological position in the warm epeiric seas during the later part of the Lower Car- 
boniferous. But their opportunity came too late for them to make much of it. 
Although individuals became so extraordinarily numerous that they cause some late 
Mississippian strata to resemble conglomerates, and although the major genus, 
Pentremites, has hundreds of species, yet there is really no great variety among the 
blastoids. After their brief day of glory in late Mississippian times, they returned to 
their former insignificance. Indeed, they became extinct in all but a few regions. 

Oldest and most important, and, to some, the most fascinating of the echinoderms, 
are the cystids. Among them the scientifically imaginative can see the ancestors of all 

FIG. 34. Camerate crinoids from the Mississippian of Iowa. At left, a calyx 
with arms and anal tube; next, a similar crinoicl lacking arms, showing tegmen 
and anal tube, and the openings at bases of the arms through which food 
passed to the concealed mouth; third, and at right, a somewhat simpler type, 
drawn in lateral and lower (dorsal) views, showing the form of the calyx 
and method of branching of the arms. From Wachsmuth and Springer. 

FIG. 35. Three kinds of cystids. At left, Cheirocrinus, one of the Rhombifera, with erect 
arms on the summit of the calyx; next, another of the Rhombifera, with sessile food grooves; at 
right, actinal and lateral views of one of the Diploporita, with numerous pores in pairs, and sessile 
food grooves. All from Otto Jaekel. 

FIG. 36. At left, a bit of Ordovician sea floor, with four cdrioasteroids. About half natural 
size. At right, a modern brittle star, with only one of the five arms shown. Original drawing 
by the late Professor William Patten. 

FIG. 37. Three views of a Mississippian pentremite, to show the large plates 
of the calyx, the five ambulacral areas, the central position of the mouth, and 
ihe openings around it. From J. F. Pictet, Traite de paleontologie. 

FIG. 38. A branching kind of Favorites, one of the most common genera 
of tabulate corals. 


other members of the phylum; but it must be admitted that, if absolute proof of this 
theory is demanded, the record now known is not fully convincing. 

The oldest cystids (Lower Cambrian) are sac-shaped. But a sac has no particu- 
lar shape; that depends upon what is in it, and whether or not it is full. Such a 
description fits the early cystids to a nicety. They had no particular shape, no sym- 
mtry. They can be called echinoderms only because their skeletons were made up 
of plates and because each had a water-vascular system. This last is not expressed in 
ambulacral areas radiating from the mouth, for true cystids had no ambulacral 
grooves, and the position of the mouth of many is difficult to locate. The presence 
of a water-vascular system is shown among the oldest and most primitive of them 
by numerous perforations in every plate (Fig. 35, at right). In the earliest it was not 
confined to particular areas but penetrated the whole skeleton. 

Evolution amongst the cystids is best expressed by the statement of certain tenden- 
cies. First, there was a tendency toward the reduction of the number of plates. Second, 
there was a tendency toward constriction of the sac at the attached end, to form a 
column. That the stalk of the crinoid is a restricted area of the primitive body cavity 
is indicated partly by the fact that it contains a "heart" and other organs, and partly by 
the fact that the distal columnals of some of the Paleozoic forms are made up of five, 
ten, or more, radially arranged pieces. Third, there was a tendency toward an 
orderly arrangement of the plates, beginning at the region of attachment and spread- 
ing toward the free (actinal) end. Fourth, the food-gathering area, originally a 
simple mouth, tended to spread. It did so in two ways: either by the production of 
food grooves along arms which are outgrowths or constrictions of the upper end of the 
original sac, or by the spreading of sessile food grooves over the surfaces of the 
plates (Fig. 35). Fifth, the number of openings for the water-vascular system tended 
toward reduction. In the later cystids (Rhombifera) they are confined to particular 
areas, where they are systematically arranged (Fig. 35, two specimens at left). There 
was originally no connection between the food grooves and the water-vascular areas. 
One small group of cystids lost the latter system entirely. Only a few of the older 
crinoids show external traces of it. In the starfish and blastoids, on the other hand, 
it is conspicuously developed, and combined with the food grooves in the ambulacral 
areas. Sea urchins have a highly developed water-vascular system, but no food grooves, 
although the tube feet occupy the ambulacral areas. Holothurians, the most specialized 
echinoderms, have an echinoid type of water-vascular apparatus. 

Various cystids show various combinations of the results of the tendencies listed 
above. One line, characterized by loss of the water-vascular system, great reductions in 
the number of plates, and the development of food grooves on the arms, led to the 
crinoids. Another, in which there was less reduction of plates, but a restriction of 
pore-bearing ones to five particular areas, seems to have started the sea urchins in 
Ordovician times. Still another group, the edrioasteroids (Fig. 36, at left), in which 


food grooves and water-vascular systems were combined, may have been ancestral to 
the starfish, and they in turn to the brittle stars (Fig. 36, at right) . The sea cucumbers, 
or holothurians, with much reduced skeleton, may have come from the same stock, 
although many believe that they are more closely related to the primitive sea urchins. 
Although the early blastoids remained attached and actinal side up, they must have 
had the same ancestors as the starfish. 

All of which is more or less speculation. How fully justified, further research 
alone can tell. As for the cystids themselves, they appeared first in the Lower Cam- 
brian, where they are rare. Rather specialized, flattened, possibly motile forms are 
found in Mid-Cambrian deposits. Unfortunately for Dr. F. A. Bather's ideas, out- 
lined above, of the origin of the various other echinoderms from the cystids, the really 
primitive members of the group have so far been found only in Ordovician rocks. 
That, however, is no real objection to the acceptance of the theories, for there are 
numerous cases where simple, primitive animals have continued to live long after 
their more specialized derivatives have disappeared. The cystids seem to have reached 
their culmination in abundance and variety during the Middle Ordovician. Except 
at a few localities they are extremely rare in Silurian and Lower Devonian strata, and 
the last of them are found in rocks of Mid-Devonian age. 

A great deal is known about the history of the sea urchins and something about 
that of the sea cucumbers, but there is no space to record it here. The echinoderms are 
so numerous, so beautiful, and so scientifically important that a whole book would do 
them less than justice. 


You strange, astonish'd-looking, angle-faced, 
Dreary-mouthed, gaping wretches of the sea, 
Gulping salt water everlastingly, 

Cold-blooded, though with red your blood be graced, 
And mute, though dwellers in the roaring waste . . . 

Leigh Hunt, "To a Fish" 

The animals o the earlier Paleozoic had no backbones. It is believed that verte- 
brates were derived from some sort of invertebrate, but the problem of their origin 
still baffles zoologists and paleontologists. Zoologists have attacked it from the stand- 
point of comparative morphology; that is, they have studied all sorts of vertebrates, 
with the intention of learning which attributes are primitive and which specialized. 
Paleontologists, in their search for missing links, have studied fossils rather than 
living animals. In Devonian strata they have found many puzzling objects which, 
although they show some of the features of vertebrates, do not fit readily into the 
modern categories of such animals. In order to understand these curious creatures 
qne must first investigate modern fish and learn something of their fundamental 

Most fish have a cartilaginous or bony skeleton consisting of three parts. One of 
these, including the skull and backbone, is the axial skeleton. The backbone is seg- 
mented, the parts represented by numerous biconcave vertebrae, each separated from 
its neighbor by cartilage. The study of fish embryos reveals that no backbone is 
present in young stages, its place being filled by a continuous unsegmented rod known 
as the notochord. This organ is not formed of strong tissue but is made up of watery 
cells enclosed in a tubular sheath; it may be likened to a rubber container surcharged 
with water. Although not in itself strong, it gives the embryo a longitudinal support 
around which the vertebrae develop, growing inward from the outside, until in the 
adult they subdivide it, enclosing its remnants between their concave ends. The an- 
terior end of the notochord is beneath the brain, a part of the base of the skull develop- 
ing around it. Since, in the ontogeny of the individual, the notochord precedes the 
backbone, it is probably more primitive than the vertebral column, whence the name 
Chordata, rather than Vertebrata. The axial support is, therefore, the distinguishing 
feature of the phylum. Two sections of the skeleton, the visceral and appendicular 
portions, are pendent from it. 

The visceral skeleton is composed of from seven to nine inverted arches, although 


some sharks have vestiges of more. The first arch is double, forming the upper and 
lower jaws. Next is the hyoid arch, a support located in the throat. Beneath the 
posterior part of the skull are from five to seven similar arches which sustain the gills. 

The appendicular portion consists of the limbs and their supports. In modern 
fish this system has two parts, the anterior pectoral, and the posterior pelvic paired 
fins, with their girdles. As will be seen, however, some of the ancient ones possessed 
more than two pairs. The pectoral girdle of the bony fish is made up of several bones 
and is attached to the back or sides of the skull. The pelvic girdle, on the other hand, 
is poorly developed; absent, indeed, from modern ones. It has no connection with the 
backbone, even in sharks, where it serves to support the relatively small posterior 
pair of fins. These general facts about the nature of the skeleton suffice for current 
purposes, so we may now turn to a more general consideration of the fish world. 

At the present day there appear to be three principal groups of fish, most easily 
distinguished by the nature of their scales. The shark (Fig. 39) or skate has no con- 
tinuous external skeleton, but numerous individual, more or less thornlike scales 
embedded in a tough skin. In section each scale shows an internal pulp cavity sur- 
rounded by a hard compact substance like the dentine of our teeth, traversed by 
numerous branching but minute canals radiating from the pulp cavity. The external 
layer consists of extremely hard, dense enamel, the hardest of all organic tissues. 
Under the microscope this substance appears uniform in texture, although it may be 
traversed by extensions of the canals of the dentine, which do not, however, branch 
in the enamel. Such scales, called placoid, may or may not be mounted on a bony 
base. Since they are found only in sharks, skates, and their allies, they characterize a 
subdivision of the fishes, known as the elasmobranchs. Another type of scale is found 
in the ganoids (Fig. 39). These fish, represented at the present by the gar pikes and 
sturgeons, possess a covering of scales, diamond-shaped or, more rarely, circular, 
overlapping one another like tiles and in most cases interlocking by toothlike 
processes. When examined in section each scale shows a bony substratum coated 
externally with a substance known as ganoine, enamel-like in structure. Hence the 
ganoids are often spoken of as the enameled-scaled fish. It should be noted, however, 
that there is no pulp cavity or true enamel. The bony fish, teleosts, have thin flexible 
scales or none at all. Their scales are referred to as cycloid or ctenoid, according to 
whether their margins are smooth or possess comblike projections. 

The skeleton of the elasmobranch consists almost entirely of cartilage, or of carti- 
lage impregnated with calcium carbonate; that of the ganoid is also largely cartilagi- 
nous, especially among the more ancient ones. As the evolution of the ganoids is 
traced, however, it is found that with the passage of time the skeleton became more 
and more ossified, whereas the scales became thinner and thinner, the change marking 
a gradual transition from the ganoids into the teleosts, whose skeleton is completely 
bony. Most modern writers consider the ganoids as a subdivision of the teleosts. 

FIG. 39. Above, a modern shark, showing position of mouth, gill slits, 
and heteroccrcal tail fin. Water color by J. Burkhardt, 1864. In middle, a 
modern ganoid, showing scales and heterocercal tail fin. Wash drawing by 
Jos. Dinkel, 1834. Below, skeleton of a modern bony fish (telcost), with 
homocercal tail fin. All from drawings, of the time of Louis Agassi/., in 
the Museum of Comparative Zoology, Harvard University. 

FIG. 40. Skull and pectoral fins of a Partheus, a large Cretaceous teleost. 
This skull, eleven inches high, is in the Museum of Comparative Zoology, 
Harvard University. It was found in northern Queensland. 

FIG. 41. Phareodus, a large hocene Lelcust hum Wyoming. Length about 
twenty-two inches. Related fish are found in tropical fresh waters today. 
From a specimen in the Museum of Comparative Zoology, Harvard University. 


Another variation in fish is in the nature of the tail fin. Three principal types 
exist. In the first the fin extends completely around the posterior end of the fish, 
equal parts being above and below the backbone. This type, the diphyceral, was 
formerly believed to be the most primitive, but recent studies do not confirm this 
hypothesis. It is found, however, in some of the most ancient fishes, as well as in 
modern ones which have a specialized, elongate, eellike form. It is symmetrical, 
but another common sort, the heterocercal, is unsymmetrical, the backbone extending 
into its upper lobe, whereas the lower one, which may be the larger, receives support 
only from the fin rays (Fig. 39, at top and middle). Practically all sharks and skates, 
and most of the ganoids, display the heterocercal tail fin. Finally, there is the homo- 
cereal type, exhibited by the teleosts, in which the fin lies mostly behind the posterior 
end of the vertebral column. These fins are symmetrical, many with a reentrant angle 
at the posterior end, as seen in profile (Fig. 39, at bottom). 

From these descriptions it seems obvious that specimens of the three principal 
sorts of fish should be easily recognized as fossils. Since the only really hard parts of 
the elasmobranchs are the teeth, scales, and fin spines, few of them are preserved in 
their entirety, and they are known chiefly from fragments. The ganoids, on the other 
hand, having a relatively strong external covering, may be completely preserved, but 
reveal little of the internal skeleton. Although the thin scales of the teleosts may be 
lost, their bony skeletons are readily recognizable (Figs. 40, 41). 

The vertebrates are the most familiar of the chordates, but certain living as well 
as extinct animals have no backbones and yet show vertebrate rather than invertebrate 
characteristics. We must, therefore, list those characteristics in which the Vertebrata 
appear to differ from the Invertebrata. These are: a backbone developed about or 
replacing a notochord; a tubular central nervous system which lies above the vertebral 
column, which is in turn above the alimentary canal; an anterior expansion of the 
nerve cord consisting of a tripartite brain; a variable number of visceral arches with 
gill slits between them; internal supports for the paired appendages; and bone, 
dentine, and enamel. 

It may seem from the nature of their work that paleontologists would be little 
concerned with brains. However, although they are themselves soft tissues, brains 
are surrounded by cartilaginous or bony structures with cavities which, on the de- 
struction of their contents, may be filled with mud or other substances. Such natural 
casts of the cranial cavities occur among fossils in rocks as ancient as the Silurian, 
and their study has in recent years enabled us to understand some previously obscure 
fossils. Fundamentally, the brain is a hollow object, an anterior enlargement of the 
tubular nervous system, divided into three parts. The anterior or forebrain has at 
the front the olfactory lobes, behind which are located the cerebral lobes. The mid- 
brain is connected with the power of vision and bears the optic lobes. At the anterior 
end of the hindbrain is the cerebellum. Behind it is the medulla oblongata, which 


tapers backward into the main nervous cord. The tripartite brain really, then, has 
five principal areas, although at its first formation only three parts are discernible. 

The paleontologist must be able to recognize bony tissue, which has a structure 
readily identifiable when thin slices are studied under the microscope. A section cut 
transversely to the longer axis of a bone shows circular openings which are known as 
Haversian canals. These are surrounded by concentrically arranged lamellae, among 
which are situated numerous narrow openings or lacunae; from the latter radiate 
minute tubular canals crossing the lamellae transversely, producing a more or less 
checkered appearance. Such structures, although not entirely unknown among the 
invertebrates, are characteristic of true bones, and enable the paleontologist to decide 
whether a doubtful specimen belongs to the chordate phylum. 

The oldest fossils which have been identified as allied to the vertebrates are 
enigmatic creatures with strong external and no internal skeletons. Although in this 
respect they resemble invertebrates, the posterior part of the body is fishlike, covered 
with scales. Because of the massive plates with which the anterior part of many is 
protected, these organisms have been grouped together under the name of ostra- 
coderms, or crusted-skinned animals. Although they have been known and studied 
by paleontologists and zoologists for many years, only recently have their relationships 
begun to be understood. There seem to be five types> which will be described briefly, 
with an attempt to show their relationships. 

The most ancient well-preserved ostracoderms are found in mid-Upper Silurian 
strata in northern Europe and in Pennsylvania. They are small forms known as 
Pteraspis and Paleaspls\ for our purposes they may be referred to as the pteraspids. 
Ptcraspis is the better known genus (Fig. 42). It appears to have been a slender 
animal, from three to seven or eight inches in length, the anterior half encased in an 
oval buckler composed of a few thick plates, whereas the posterior part bears thin 
scales, rarely preserved. The eyes are lateral, and the mouth, only recently discovered, 
is a transverse slit just beneath the anterior margin, bordered in front by a plate bearing 
a large median and two lateral processes which may be interpreted as having had 
the function of teeth. Behind the mouth are sixteen elongate scales which may have 
been capable of movement against the anterior plate, thus producing a sort of chewing 
apparatus. These are, however, not teeth or jaws in the true sense. No internal 
skeleton and no appendages have been found. 

Many specimens of a much larger ostracoderm, Drepanaspis (Fig. 42, at right), 
have been found in roofing slates of Lower Devonian age near Gemunden, Ger- 
many. Drepanaspis has a broad, flattened shield with large median and lateral plates, 
connected by a mosaic of smaller ones. The posterior part, which is mostly tail fin, 
is short, covered with overlapping scales. The transverse slit-like mouth is bordered 
by numerous plates covered with small denticles. The ocular orbits are small, widely 
separated, situated near the anterior corners of the shield on the dorsal side. 


Drepanaspis is allied with the pteraspids not only by the similarity in the position 
of the principal plates of the buckler and the location of the eyes and mouth but also 
by the microscopic structure of the shell. In thin section, both show a bony layer 
overlaid by compact dentine with pulp cavities and ramifying canals. W. L. Bryant has 
recently demonstrated that the oldest known ostracoderms, described many years ago 
from material collected from an Upper Ordovician sandstone at Canyon City, Colo- 
rado, have plates with the same histological structure. Although ostracoderms occur 
in vast numbers at Canyon City, all the material so far collected is extremely frag- 

FIG. 42. At left, restorations of the dorsal and ventral surfaces of Ptcraspis, 
to show the few large plates of the anterior shield. One-third natural size. 
From E. I. White. At right, the ventral surface of Drcpanaspis, to show the 
numerous small plates between the few large ones. One-ninth natural size. 
From H. C. Stetson. 

mentary, not one complete plate having been recovered. When better specimens are 
found, they will probably prove to have a form much like that of the pteraspids. It 
should be noted that in the latter the plates of the buckler are fewer in number and 
more solid than those of the Devonian Drepanaspis^ a fact which suggests that in 
the evolution of this group there was a tendency to produce a more flexible external 
skeleton by the increase in number and decrease in size of the plates. 

A second group, called ostracoderms merely because its relationships have been 
obscure, is exemplified by the genera Thelodus and Lanarfya. High in the Silurian 
in western England is a thin layer which has long been known as the Ludlow bone- 
bed, because of its typical development in the vicinity of that picturesque town. The 
bone-bed contains few if any true bones but is full of toothlike scales. Similar scales 


are found in other localities, particularly in Upper Silurian strata on the island of 
Oesel. Most of these isolated scales are considerably water-worn; hence the various 
species of Thelodus which have been described with them as a basis are really un- 
identifiable. Mr. David Tait, the veteran collector and geologist of the Geological 
Survey of Scotland, succeeded in finding complete specimens of this genus and of the 
related Lanartya in the Upper Silurian of the Lowlands of Scotland. They were 
described first by Dr. R. H. Traquair, but recently all the old and much new material 
has been scrutinized by Mr. H. C. Stetson in the paleontological laboratory at Har- 
vard. Mr. Stetson has also been fortunate enough to discover large and fairly com- 
plete specimens of Thelodus in Silurian strata north of St. John, New Brunswick, and 
scales in the Upper Ordovician at Canyon City, Colorado. 

Even with the best of material, comparatively little can be learned about the 
anatomy of these animals. There is no buckler of large plates, the entire body being 
covered with small, roughly cubical scales, some of which appear to be connected to 
those adjacent by spinelike processes. The body, which was from two or three inches 
to a foot or more in length, probably had much the shape of a modern catfish, being 
somewhat blunt-headed. Although there is no evidence of true fins, unsupported 
lateral expansions back of the middle, which may have served as balancing organs, 
suggest skatelike habits of locomotion. The eyes are anterior and lateral, far apart 
as in Drepanaspis, and the mouth seems to be a transverse slit near the anterior ventral 
margin of the head. It is bordered by somewhat smaller, more thornlike scales than 
those covering the remainder of the body. Whether they were toothlike in function 
remains to be determined. At any rate, the creatures were agnathous, jawless. 

The most important fact about these animals has been learned from thin sections 
of the scales. The histological structure is exactly that of a tooth. Each has a large 
pulp cavity surrounded by dentine, the whole capped by a layer of enamel. The 
scales are therefore of the true placoid type, which would indicate that Thelodus 
and Lanarfya belong to the elasmobranchs. The pteraspids and Drepanaspis are so 
closely related by 'their general features and the microscopic structure of their plates 
that they must be placed in the same group. Since Thelodus is now known to have 
lived in Ordovician times, it is probable that the forms now known represent at least 
two genetic lines, descending from an as yet unknown Cambrian or Ordovician 
forebear. There is a suggestion that Thelodus may have been a member of the line 
which gave rise to the true sharks; the other lineage (pteraspids) probably died out 
during the Devonian. 

The third group to which attention should be directed is that of the cephalaspids, 
long known but only recently understood. The oldest occur in mid-Upper Silurian 
in Shropshire, in Scotland, and on the island of Oesel. Scotland and Spitzbergen have 
supplied most of the specimens of Devonian age, although some have been found in 
Canada. Until the last decade only a few good specimens were known, but Nor- 


wegian expeditions to Spitzbergen brought back great numbers of them, which have 
been cleverly prepared and studied by Dr. E. A. Stensio of Stockholm, with results 
of far-reaching importance. 

The head shield of a cephalaspid is shaped like the curved blade of a halberd, 
with backward directed lateral projections (Fig. 43). It protected only the head and 
anterior part of the body, the remainder being fishlike and covered with long, narrow 
scales. The head shield seems at first sight to be composed of numerous polygonal 
plates joined together at the edges, this appearance being emphasized in some by 
the presence of small knobs, one in the center of each polygon (Fig. 45, at left). 
Investigated in thin section, this condition is traceable to the presence of canals which 
radiate from centers within the test. This suggests the structure seen in plates of 
echinoderms, particularly cystoids and crinoids. There is no evidence, however, of 
suture lines bounding the plates of the greater part of the shield. There are three 
areas covered by small, thin plates. Two of these are lateral, whereas the third, on 

FIG. 43. A restoration of Ccphalaspis, by E. A. Stensio. Reproduced by 
permission of the Trustees of the British Museum (Natural History). 

the median line, is called the "postorbital valley" because it appears in the fossils 
as a depressed area behind the eyes. The orbits are large, close together, near the 
middle of the top of the shield. On the median line, just in front of the eyes, is a 
small opening; behind it, and between the eyes, is another smaller one. These open- 
ings were explained some years ago when Professor Carl Wiman of Upsala obtained 
from Oesel a weathered specimen which exhibited a cast of the brain. It is a typical 
primitive tripartite brain, the anterior end of which lay beneath the anterior of the 
two openings just mentioned. Since this is the location of the olfactory lobes, the 
opening is to be interpreted as that of the nose. The posterior opening is above the 
hind part of the forebrain, and corresponds to the position of the epiphysis leading to 
the median or pineal eye. Dr. Stensio, by making numerous skillful dissections of 
specimens from Spitzbergen, has succeeded in working out the central nervous system 
in great detail (Fig. 45, at right). He has not only developed the endocranial casts 
but has identified the ten pairs of cranial nerves so characteristic of the vertebrate. 
There is, then, no possible doubt that the cephalaspids are allied to the vertebrates, 
in spite of the absence of appendages, jaws, and internal skeleton. 

It will be noted that there are few characteristics common to the cephalaspids 


and the two other groups described above. Although bone is present in the test, there 
is nothing so far known which suggests the presence of such tissues as true dentine 
or enamel in their shells. 

The fourth group is the one which Traquair named the anaspids, because, like 
Thelodus, they lack a buckler. In all other respects, however, they are totally unlike 
that primitive elasmobranch. First found about thirty years ago in the Upper Silurian 
(Downtonian) of Scotland, these fossils were at the time entirely misunderstood. 
Our real knowledge of them dates from Professor Johan Kiaer's monographic de- 
scription (1925) of the specimens which he collected in strata of the same age in the 
Ringrike district of Norway. Stetson has increased our information by his studies 
on material from the original localities in Scotland. 

The anaspids are unique among the ostracoderms in that they have a really 
fishlike shape (Fig. 49). All are small, from one to eight inches in length, with a 
body covered by numerous elongate scales. The tail fin is unsymmetrical, the main 
portion of the body extending, curiously enough, into the lower, rather than the upper, 
lobe. This is the reverse of the usual heterocercal tail, and Kiaer suggested that it 
be called hypocercal. Such a condition is known in the remainder of the animal king- 
dom only among the ichthyosaurs and marine crocodiles, a good example of the truth 
that similarity of structure does not necessarily indicate relationship. The head is 
remarkably short, there is no neck, and the arrangement of the scales suggests that 
of the bands of muscle in modern fish (Fig. 50). They are directed forward on the 
back, backward on the sides, and forward again on the belly. A single pair of lateral 
spines indicates the presence of pectoral fins; above these on either side is a row of 
circular openings probably connected with the organs of respiration. The mouth is 
terminal, surrounded by small plates variously arranged in the different species, but 
without true jaws or teeth. The eyes are far forward, close together, high on the 
head, and there are two small openings on the median line which may, from analogy 
with those of the cephalaspids, be identified as the narial and pineal apertures. Be- 
hind the latter is 'an area of small plates corresponding to the "postorbital valley" 
of the cephalaspids. It is, therefore, a fair inference that the anaspids had a vertebrate 
type of brain, and that they are more closely allied to the cephalaspids than to any 
other group. They are discussed at greater length in connection with the theories 
of the origin of the vertebrates. 

The last and probably the most specialized group of ostracoderms is that known 
as the Antiarcha. The best known of these are Pterichthys, found chiefly in the 
Middle and Upper Devonian red sandstone of Scotland, and Bothriolepis, a rather 
widely distributed fossil in strata of the same age in Quebec, eastern New York, 
and Pennsylvania (Fig. 44). These are the original and real "ostracoderms," for the 
anterior part of the body is encased in a shield of thick, solid plates. It is rather 
amusing to find that the latest workers exclude them from that group. They differ 

FIG. 44. Dorsal and ventral views of Bothriolepis, as reconstructed by Pro- 
fessor William Patten, From Patten's The Evolution of the Vertebrates and 
Their Kin, by permissioi of P. Blakistons Son and Company, Inc. 

FIG. 45. At left, the head-shield of the Silurian cephalaspid, Attchenaspis. Note the narial 
opening in front of orbits, the pineal opening between them, and the postorbital series of small 
plates. From J. Kiaer. In middle, cranial roof of an anaspid, showing the same features. From 
J. Kiaer. At right, a dissection of a Devonian cephalaspid, showing casts of the cavities occupied 
by the brain, principal nerves, and other organs of the head. From E. A. Stensio. 


from all others in having a head shield which is movably connected with a boxlike 
case covering the anterior part of the body, and in possessing a pair of jointed append- 
ages. The posterior part of the body is fishlike, with a dorsal fin and a heterocercal 
tail. This part of the animal is protected only by the thinnest scales, and would 
hardly be known except for the discoveries made by Professor William Patten at 
Scaumenac on Chaleur Bay in Quebec. The eyes occupy opposite ends of a trans- 
versely placed opening in the top of the head, separated by a plate not ankylosed to 
the shield and hence rarely preserved. The mouth is ventral, near the margin, bounded 
laterally by three pairs of plates which may be variously interpreted. Patten, who 
discovered and described them, believed that they represented jaws which moved 
inward toward the median line, but it has more recently been suggested that they may 
have moved up and down. The external covering has the histological structure of bone. 

The specimens from Scaumenac are of particular interest because they are so 
well preserved that a few carbonized remains of the internal organs are found, re- 
vealing something of the anatomical structure. Sections show what seem to have 
been gills and a stomach, but no remains whatever of vertebrae, notochordal sheath, 
gill arches, or any appendicular support. Although this is negative evidence, it strongly 
supports the view that there were not even rudiments of an internal skeleton in the 

To summarize what has been said above, the various fossils which have been 
referred to the ostracoderms represent five distinct groups. Two of these, typified 
by Thelodus and Pteraspis, are shown by the microscopic structure of the scales 
and plates to be allied to the shark-skate group, and should be transferred to the 
elasmobranchs. The cephalaspids and anaspids are probably related to each other. 
The form of the brain shows that they are true chordates. In the opinion of the 
writer, the anaspids are the more primitive forms, possibly ancestral to the ganoids, 
whereas the cephalaspids were somewhat specialized bottom-dwellers which left no 
descendants. Dr. Stensio's argument that they were the ancestors of the modern 
cyclostomes is of interest but cannot be discussed in this book without entering into 
too many technicalities. The last group, that of Bothriolepis and its allies, shows no 
relationship to the others, but occupies an isolated position. There is considerable 
evidence to show that it is related to the arthrodires, a group of chordates not closely 
allied to any other known animals. 

Although they are not classified with the ostracoderms, some of the arthrodires 
are so ostracoderm-like that they may be described in connection with them. One of 
their principal characteristics is that their armor consists of two portions: a buckler 
over the head, and another over the neck and anterior part of the trunk. These articu- 
late with one another by a hingelike joint, just as the two shields of the Antiarcha do. 
The best known arthrodires are the small Coccosteus (Fig. 46) of the Mid-Devonian 
of Scotland and the huge Dinichthys and Titanichthys of the Upper Devonian of 


Ohio. The Scottish animal is most like the ostracoderms, for the anterior part is 
covered by a shield whose sculptured plates show it to have been largely external. 
The posterior part of the body, however, not only is fishlike in shape, with a hetero- 
cercal tail, but has a series of neural and hemal arches throughout its length which 
enclose a notochord and form the rudiments of a backbone. 

Remains of Dinichthys are rather common in a black shale in the vicinity of 
Cleveland, Ohio (Fig. 47). Although its large platelike bones have been known for 
a long time, recent collections and studies have added much to our knowledge of 

FIG. 46. A reconstruction of the Mid-Devonian arthrodire, Coccosteus. 
Note the space for a notochord between the neural and haemal spines. From 
A. Heintz. 

FIG. 47. The skull of the huge arthrodire, Dinichthys, to show the "teeth" 
and, diagrammatical ly, some of the principal muscles. From A. Heintz. 

this extraordinary carnivore. Some of the shales in which the bones occur are so 
near to the center of Cleveland that they were in danger of being completely lost as 
collecting places through the expansion of the city. At the crucial moment, however, 
public-spirited citizens raised sufficient money to allow their exploration by means 
of a steam shovel. As a result of this modern and thorough instrument of search, 
new material has been obtained which allows an accurate restoration of these extinct 
animals. A great fish, with a skull nearly three feet across, and a body ten or a dozen 
feet long, Dinichthys was in marked contrast to the little Coccosteus. Unlike those of 
the latter, the shields were internal, entirely covered by flesh and scaleless skin. The 
jaws were fearful, armed with trenchant teeth more efficient than those of most sharks. 
Yet in reality Dinichthys had no teeth. It bit by working the whole top of the head 
instead of snapping shut its lower jaws. The latter may have been held rigidly 
while one set of muscles lifted the head shield, and another pulled it down. Sharp- 


bladed projections of the bones of both upper and lower jaws served as teeth. Although 
they are toothlike in form, thin sections show that these projections are nothing more 
than dense bone, without dentine or enamel. 

The arthrodires are the only creatures in the whole animal kingdom which 
actually use their jawbones for biting. They were not senile and toothless grandsires, 
mutpbling their food, but were capable of quickly slicing in twain any of their 
contemporaries. Dinichthys and Titanichthys, the latter possibly fifteen to twenty 
feet long, were truly the rulers of the late Devonian or early Mississippian seas. Yet 
this group suddenly disappeared, at the zenith of its powers. The too rich and power- 
ful are always on the verge of extinction. 


Causa latet: vis est notissima. 

''The cause is hidden, but the result is known." 

Ovid, Metamorphoses, iv, 287 

It was pointed out in the earlier pages of the preceding chapter that there are 
many fundamental differences between the two great divisions of the animal kingdom, 
the Chordata and the Invertebrata. Those who try to trace the path along which 
evolution has proceeded find in this region only the faintest trail. As the discussion 
of the ostracoderms has shown, these fishlike animals were not full-fledged verte- 
brates. Although in some respects they stand between the vertebrates and the in- 
vertebrates, nevertheless they give few clues to the particular phylum from which 
the higher group sprang. The problem of discovering the connecting links belongs 
both to the zoologist and paleontologist. It is still far from solution, but some progress 
is being made. In this place all that can be done is to present the current theories, with 
some of the evidence for and against them. 

The one which at present has widest acceptance is derived entirely from the study 
of certain living animals, all, unfortunately, entirely without hard parts, and so not 
likely to be preserved as fossils. The line proceeds downward from primitive fish 
through Amphioxus, thence to the larval tunicate, then to Balano gloss us, and finally 
to some unidentified echinoderm. Since the presence of a notochord is the most im- 
portant factor, this theory is commonly called the "chordate" or, because of the promi- 
nent place that animal occupies in it, the "Amphioxus" theory. 

Amphioxus, the simplest living true chordate, is a small, slender, cigar-shaped 
animal, from one to two inches in length. It has a worldwide distribution in warm 
shallow waters, where it lives partially buried in the sand, the mouth protruding. 
It is nevertheless capable of swimming, even of active movements. It has no distinct 
head; the mouth is ventral, a little behind the anterior end, with paired gill slits back 
of it. Although there are no real fins, a median dorsal fold and a pair of ventro-lateral 
ones suggest rudiments of such organs. The notochord extends from the anterior to 
the posterior end and constitutes the entire axial support, for the creature has no skele- 
ton. Above the notochord is a tubular nerve cord, as in the vertebrates; below it, a 
straight alimentary canal. A conspicuous feature of the animal is the regular seg- 
mentation of the muscles, which may be plainly seen because it is translucent. They 
have the form of V-shaped segments, with the angle pointing forward. The segmenta- 


tion, having an outward expression, is much more obvious than that of the higher 
vertebrates, in which it is shown only by internal characteristics, such as the repetition 
of vertebrae in the spinal column, nervous ganglia, and the like. The muscles of a 
fish (Fig. 50), however, have the same arrangement as those of Amphioxus^ as one 
may observe on any Friday. . l 

^ Although it possesses only a few characteristics of the higher vertebrates, those it 
has are so important that no one now denies Amphioxus a place in the same phylum 
with them. It is generally believed that the fish arose from an animal much like it, 
even though the connection cannot be traced. Furthermore, the search for its an- 
cestors has not yet met with success. The one contact between it and the invertebrates 
is found in the fact that the larva is ciliated, swimming by means of hair-shaped 
outgrowths. The possession of a ciliated larva is common among the invertebrates but 
entirely unknown among the craniate vertebrates; however, this is only a vague con- 
necting link with the former" subkingdom. It is possible that Amphioxus may have 
been derived from some member of the Hemichordata, a group in some respects more 
primitive. These animals, the tunicates or sea squirts, are commonly included as links 
in the chain of this theory because during their larval stages segmented muscles, gill 
slits, and a notochord are present. So far as they are now understood, they really throw 
no light on the problem. 

There is one group of living animals which may represent a simpler type of 
chordate than the larva of the tunicates. The best-known member is Balanoglossus, 
a wormlike marine creature which lives in fragile tubes in the region between high 
and low tides (Fig. 48). At the anterior end of the "worm" is a long, sensitive pro- 
boscis, capable of great extension and contraction. Behind the collar is the trunk or 
body proper, on the anterior part of which are numerous pairs there may be as 
many as fifty of openings or gill slits, much like those of Amphioxus. In the in- 
terior of the proboscis is an axial rodlike structure which arises as an outgrowth from 
the alimentary canal above the mouth. Because of its position, its origin, and the 
nature of its cells, this rod has been considered a notochord. In another genus, Harri- 
mania, there is a similar but much larger structure, strikingly like a true notochord. 
According to S. F. Harmer, Balanoglossus is "provided with a dorsal, tubular, central 
nervous system, and although it does not extend beyond the limits of the collar," it 
shows noteworthy resemblances to that of vertebrates. This protochordate is therefore 
considered the simplest existing representative of the group from which the vertebrates 
were derived. 

Even Balanoglossus, simple as it is, does not serve to show positively from what 
phylum the chordates were derived. A suggestion, however, is furnished by the form 
of its larva. Students of zoology have learned that the very young, free-swimming 
larvae of all members of the same group are much alike, and have therefore come 
to depend to some extent upon the young as indicators of relationship. In this case, 


if one were to judge from the form of the adult, Balanoglossus would seem to be 
closely allied to the worms. The free-swimming larva, however, is unlike that of the 
worms but rather resembles that of the echinoderms. In fact, it is so much like the 
latter that when first discovered it was supposed to be the young of some member 
of the latter group. Although it is not possible to go into detail about these larvae, 
the diagrams (Fig. 48) will illustrate the resemblances between those of the echino- 
derms and Balanoglossus, and the difference from those of the worms. 

In addition to the similarity of the larvae, there are other ways in which Balano- 
glossus appears to be allied to the echinoderms. It has a system for the circulation of 
water comparable to their water-vascular system, and H. H. Wilder has found that 
its five body cavities are probably represented in the echinoderm larvae. No proof 
of connection between the echinoderms and the chordates is yet forthcoming, but we 


FIG. 48. A, trochophore larva of an annelid, compared with (B) the 
free-swimming larva of Balanoglossus , and (C) that of a crinoid. A, 
after B. Hatschek; B, after W. H. Bateson; C, after Goette and Thompson. 
At right, a diagrammatic representation of Balanoglossus, simplified after 
W. H. Bateson. 

have the interesting suggestion that a bilaterally symmetrical ancestor gave rise to 
the radially symmetrical, lowly echinoderm and the progressive and finally dominant 
vertebrate. Here is a place where paleontology should come forward with definite 
evidence, but it has as yet failed to do so. 

Another idea which has been more or less popular among zoologists is that em- 
bodied in the "annelid theory." The fundamentally segmented nature of the verte- 
brates naturally suggested the idea that they arose from the annelids, a view not yet 
entirely abandoned. The chief ground for the hypothesis is the similarity which exists 
between embryonic vertebrates and adult annelids. For example, the vertebrate em- 
bryo shows distinct repetition of parts, and even in the adult the vascular system, the 
nervous chain with its paired ganglia, and the general relationship of the principal 
organs to one another show considerable similarities to those of the annelids. Some 
authors have maintained that the bundle of fibers which in some annelids supports 
the nervous chain is comparable to a notochord, although histologically it is of en- 
tirely different structure. Moreover, both Chordata and Annelida have their internal 


organs in a body cavity, the coelom, and there is agreement in the structure of the 
kidney tubes, or nephridia. 

The most obvious difficulty encountered in .trying to derive a vertebrate from 
the segmented worm is the general reversal of all parts of the body that is necessary 
to make the change. The anterior ganglion of the nervous system of the annelid is 
abcwe the mouth, but the two cords which are the principal links in the chain pass 
around the anterior end of the alimentary canal, continuing beneath it along the 
ventral part of the body. The main blood vessel lies above the gut, with another be- 
neath it, the blood flowing forward in the dorsal and backward in the lower one. To 
transform the annelid into a vertebrate it is necessary to reverse the dorsal and ventral 
sides, abandon the old mouth, produce a new one on the opposite side, close the old 
terminal anus and form another further forward, change the solid to a tubular nerv- 
ous system, and develop a notochord between the central nervous system and the 
alimentary canal. Taken all at one time, this seems difficult of accomplishment, but 
we are assured that such things are physiologically possible. 

The theories which have most interested paleontologists are naturally those in 
which fossils play a part. Two eminent zoologists and investigators, Professors W. H. 
Gaskell and William Patten, have attempted to connect the vertebrates with the 
arthropods. Such suggestions may be called, for convenience, "arthropod" theories. 
Both men brought to the support of their causes a wealth of learning, labor, and in- 
genuity which entitles them to greater consideration than is commonly accorded. 
Since the theories are similar, although by no means identical, the discussion of one 
will suffice. The writer has chosen Dr. Patten's, partly because its exposition is some- 
what more recent, and partly because its genial author was good enough to discuss 
it freely with him. 

The arthropod theory is, in brief, that the ostracoderms stand midway between 
invertebrates and vertebrates. It does not attempt to connect any particular ostraco- 
derm with any particular invertebrate, but takes up general rather than particular 
features of the two groups. Professor Patten was doubtless influenced, however, by 
his studies of Bothriolepis and the general resemblance of its skeleton to that of the 
arachnids of the group Eurypterida (Fig. 44). Bothriolepis resembles the latter in 
that both have an external segmented armor, jointed lateral appendages with external 
but no internal supports, and, according to Patten's interpretation, similar jaws, since 
he deduced that they moved toward each other in a horizontal plane. Neither has any 
internal skeleton. Bothriolepis has only one pair of appendages, but in the arthropods 
there is a tendency toward reduction in the number of limbs, those which are retained 
being on the anterior part of the body. Hence it is logical to infer that the one pair 
of the Antiarcha may be the sole survivors of the six pairs of the eurypterids. 

The arthropod theory resembles the annelid theory in requiring an introduction 
dc novo of a notochord and the reversal of dorsal and ventral sides, with its accom- 


paniment of the production of a new mouth and the closing of the old one. There 
is some indication of a notochord in representatives of the arthropods; for example, 
a scorpion Patten found in Ceylon has what appears to be a real notochord, even the 
histological structure being apparently the same as that of the similar rod in the 
vertebrates. That the mouth of the vertebrate is a relatively new structure is indicated 
by its rather late appearance in the embryo; moreover, the fact that it is lined with 
ectodermal tissue proves that it is an inpushing from the outside. The cloacal chamber 
is likewise lined with ectoderm, indicating a similar origin for the anus. Patten pointed 
out that the tube of the alimentary canal, where it passes through the nervous system 
in the arthropod, is subject to constriction in the higher groups, in which there is an 
increase in the size of the brain because of the bunching together of ganglia in this 
region. Thus in many insects this tube is so narrow that solid food cannot be taken, 
a circumstance that causes the animals to develop suctorial habits. The process goes 
still further in the adults of many Lepidoptera (butterflies and moths), in which the 
tube is so completely shut off that they do not feed at all after the larval stage is passed. 
There is, therefore, reason to expect the closure of the old mouth in arthropods. Many 
anatomists have pointed out that the hypophysis at the base of the vertebrate brain 
is a blind sac which may be situated where the alimentary canal formerly passed 
through the central nervous system. Patten saw in the so-called "dorsal organ" of 
some crustaceans the beginnings of the new mouth. The "dorsal organ" is a sac on 
the median line which leads downward nearly to the alimentary canal. A slight 
change, involving only an opening into the digestive tube, would enable it to function 
as a mouth. 

It is not possible to enter fully into all the evidence which Patten gathered to 
support his thesis, nor is it possible to refute his arguments. The chief reason for the 
rather general lack of acceptance of the arthropod theory is that all the animals used 
as connecting links are specialized, and most zoologists believe that simple members 
of a higher group never arise from specialized ones of a lower grade. Bothriolepis 
appears to be one of the most specialized of all the ostracoderms, and the oldest of 
its relatives, found in Lower Devonian strata, have the same structure. The euryp- 
terids themselves are complexly organized, and there is no evidence that they gave 
rise to anything but the scorpions. As has been pointed out in the previous chapter, 
the only possible descendants of the Antiarcha are the arthrodires, which are much 
too highly specialized to be considered as the ancestors of fish. It is generally believed 
that all the armored ostracoderms (pteraspids, cephalaspids, and Antiarcha) were 
slow-moving, bottom-living forms, adapted to a mud-grubbing existence. There is 
little likelihood that they could have thrown off their armor and their sluggish habits, 
to be transformed into swiftly moving fish. 

The most recent of the theories proposed by paleontologists, the anaspid theory, 
although it cannot at the present time be called complete, is founded upon observa- 


tions which do not fit in with the schemes outlined above. It therefore appears neces- 
sary to state them separately, as the writer has done for several years in his lectures, 
although they now appear in extcnso for the first time. The tentative outline is that 
the earlier scaled anaspids may possibly have been derived from bilaterally symmetrical 
Cambrian ancestors of the echinoderms. From the scaled anaspids originated the 
nesirly naked Lasanius. In it an internal skeleton appears as a series of visceral 

FIG. 49. Above, an anaspid, showing the fishlikc form, the hypocercal tail, 
and the arrangement of the slender scales of the body. From Kiaer. Below, 
the head and anterior fin-support of an acanthodian, for comparison. Note 
in both the large, anteriorly placed eye with ring of sclerotic plates, the 
irregular arrangement of scales on the head, the position of the external 
fin support (Pect Sp.), and the fact that the latter is connected with an 
internal element (Sc.). Mand. Op., Hy. Op., mandibular and hyoid regions; 
Br., Ar. I, Br. II, Br. Ill, Br. IV, branchial region; mand., mandible. From 
D. M. S. Watson. 

arches, and there is probably a cartilaginous brain case and a notochord with its sheath. 
A new exoskeleton, consisting of scales with ganoine, was gradually evolved. 

The vertebrate-like characteristics of the scaled anaspids have been pointed out 
in the previous chapter. They are: the possession of a tripartite brain belonging to a 
nervous system above the gut; a fishlike form, including a tail which was an effective 
propeller, indicating the power of rapid, fishlike movement; and a pair of spines in 
a location which suggests the presence of pectoral fins. Furthermore, the arrangement 
of the scales indicates a set of muscles like those of fishes (Fig. 50). On the other 
hand, the scaled anaspids are similar to invertebrates in the lack df an internal skeleton, 
the presence of an apparently permanently open terminal mouth, the amorphous 



structure of the scales, which contain no bone, dentine, or enamel (this may be due 
to poor preservation); and the presence of plates around the mouth and the anus. 
All these features suggest an echinodermal ancestry. The same relationship is indi- 
cated by the branchial openings. They are circular apertures perforating a plate, 
hence not like the gill slits of vertebrates but suggestive rather of the openings of the 
water-vascular system so characteristic of the cystoids. Some of the Cambrian and 
Ordovician cystoids have a linear series of such openings. It must be confessed that 
the relationship to echinoderms is suggested rather than proved. We have already 
seen that there exists in the shell of the cephalaspids, the nearest relatives of the anas- 
pids, a series of radially penetrated polygonal plates, much like those of the cystoids. 

FIG. 50. Sketch of a skinned codfish, to show the arrangement of the 

Fie. 51. At left, outline reconstruction of Lasanius, showing dorsal scales, 
hypocercal tail, and internal "branchial" skeleton. The longitudinal line and 
its branches merely suggest the arrangement of the external scales. From 
O. M. B. Bulinan. At right, diagrammatic reconstruction of the internal skele- 
ton. From H. C. Stetson. 

From the scaled anaspids the scaleless form, the Scottish Lasanius, was derived. 
It differs from the other members of the group in having so thin an external covering 
that the scales are retained only by the largest and best-preserved specimens. Enough 
of these have been found, however, to show that the outward form and structure were 
similar to those of the better-known scaled genera. Its most remarkable feature is the 
possession of eight pairs of slender rods, the explanation of which has only recently 
been attempted (Fig. 51). Mr. H. C. Stetson obtained material pertaining to this 
genus from the Downtonian strata at the original localities in Scotland, and the in- 
terpretation here presented is that worked out by him. The rods appear to have 
formed an internal structure, comparable in general form to the gill arches of a 
shark. They were arranged in pairs, on opposite sides of the body. The long, spine- 
like dorsolateral portion of each broadens at the lower end, whence a straplike branch 


projects inward and backwards toward a similar one from its mate on the other 
side. During the animal's life these ventral processes were probably joined by median 
cartilages. The broad base supports a slender spine which projects outward and some- 
what backward. The first two elements, the dorsal rod and the ventral strap, were 
doubtless beneath the skin, for they show no trace of surficial ornamentation. A 
smaj.1 portion of the broadened base and the slender outer spine, on the other hand, 
show delicate striations, which indicate that they were outside the body. 

It appears, therefore, that Lasanius had an internal skeleton of basketlike form, 
the lower elements connected across the body, the upper ones separated. In seeking 
its origin, it will be noted at once that the internal elements lie in the same position 
as the external scales of the other anaspids. They may have originated as external 
scales later drawn within the skin by the pull of the muscles, but it is more likely 
that they were formed de novo as cartilages between adjacent segments of the muscles. 
The primary function of this structure is reasonably clear. The outer spines doubtless 
supported a lateral fin below the median line on each side of the body. Whether the 
internal parts had another use, connected with gills, is more problematical. In front 
of the first of the series of rods just described there is on each side a series of six short 
acicular plates arranged in a diagonally ascending row. The base of each has a 
semicircular outline, as though it fitted above a circular opening. The plates lie in 
the position of the branchial plate of the scaled anaspids; hence it is probable that 
they indicate the location of the gill openings of Lasanius. If that be the case, as is 
the opinion of Kiaer and others who have studied the question, then a pumping 
movement produced by the up-and-down motion of the fins would have aided in 
augmenting the flow of water in at the mouth and out through the circular apertures. 
If gills or gill-like sacs were present, they would probably develop first directly opposite 
the openings, but as they enlarged they would tend, because of the forward motion 
of the animal in swimming, to push backward, and so come to be supported by the 
basket. Hence a connection between the rod structure and the initiation of gill 
arches may be inferred. 

Lasanius was probably the most specialized and certainly the most active of the 
anaspids. Its external covering of scales had been reduced to a state in which they 
were no longer an adequate support for the muscles. It is therefore probable that 
the animal had some sort of internal stiffening rod. Some specimens show traces of 
what may be interpreted as the remains of a notochordal sheath, so that the transi- 
tion from such a creature to one which could actually be called a fish is not a difficult 

In fact, transitional forms may possibly be recognized in certain fossils, the oldest 
complete specimens of which are found in the Lower Devonian of Scotland. They 
are the acanthodians (Fig. 49). Many specimens have been collected and studied in 
considerable detail by paleontologists, who have found that the oldest members of 


the group are much like anaspids in the structure of plates about the mouth, the an- 
terior position of the large eyes, and the posterior location of the narial opening. 
They differ, however, in that they have toothed jaws, which seem to correspond in 
position to the cartilages of the anterior visceral arch. The scales have the histological 
structure characteristic of the ganoids, and the tail is heterocercal. Some think that the 
Acanthodii are ganoids; others have supposed them to be sharks. Dr. D. M. S. 
Watson's recent studies confirm the supposition that they are the oldest gnathostomes 
(jaw-bearing animals), although there is still difference of opinion as to their rela- 
tionship to other fish. The most striking difference from the anaspids is that they 
have a fishlike type of scale. The amorphous nature of the exoskeleton of the scaled 
anaspids suggests that it was an inheritance from an invertebrate ancestor and was 
merely the hardened outer layer of the skin. Lasanius appears to show that this old 
skeleton was in the process of being lost. It is possible that a new type of scale might 
be formed later from the inner layer of the skin after the ganoid fashion. 

All the Acanthodii are characterized by a series of small paired fins, each sup- 
ported by a single anterior spine, which is homologous with the external lateral fin 
spine of the anaspids (Fig. 52, at right). Behind the spine the membrane was un- 
supported in the early acanthodians, but somewhat stiffened by cartilages and rays 
in the later ones. As may be seen by inspection of the accompanying figure (Fig. 52, 
at left), the arrangement of the elements in the pectoral fin of Acanthodes suggests 
that of a crossopterygian (Fig. 54). In passing, it is interesting to note that these 
skeletal structures arose between the membranes of a fin supported at its anterior 
margin and hence within the area of least motion. The fact that the pelvic fins of 
nearly all the species are in front of the mid-length suggests the gradual posterior 
shift of fin spines arranged as those of Lasanius were. Of particular interest in this 
connection are the acanthodians of the genera Euthacanthus and Climatius, from 
the Lower Old Red Sandstone of Scotland. They have three or, in some cases, four 
pairs of spines, without internal supports, between the pectoral and pelvic fins. These 
genera, therefore, retain six of the eight pairs of external spines of Lasanius. The 
shoulder girdle consists of a dorsal rod with a ventral straplike piece on each side 
and carries the fin spine externally, exactly as in Lasanius. 

Some of the acanthodians were definitely notochordal, for specimens show dorsal 
and haemal arches. Watson states that some of the later forms have rudimentary 
centra, but they are too poorly preserved to be understood. In fact, bony centra 
capable of preservation appeared in fish much later than in amphibians and reptiles, 
a natural result of their aquatic habitat. Few seem to realize that the trend in the 
evolution of fish was not toward a "higher" animal, a terrestrial creature, but toward 
a better and better adaptation to their own environment. 

The acanthodians, therefore, show many characteristics which would indicate 
that they were derived from the anaspids. Most of them are too specialized to be 


placed in an ancestral position to the ganoids, but it may be that there was some animal 
with acanthodian characteristics which formed a connecting link. At any rate, we 
get from them and Lasanius a hint as to the origin of the internal visceral arches and 
the mode of reduction of longitudinal fins with numerous supports through a multiple 
finned stage to the two pairs which survive in later fish. Even the acanthodians had 
lost the internal supports of the spines between the pectoral and pelvic fins. Watson 
does not believe that early acanthodians were ancestral to early crossopterygian 
ganoids. The writer, being no specialist and therefore interested only in general 
conclusions, does. Yet Watson himself says, in discussing gill arches in a paper pub- 
lished late in 1937: "It is evident that ultimately a Teleostome-like arrangement of 
V-shaped gill arches, separated by extremely long gill slits and bearing long gill 

FIG. 52. At left, restoration of an acanthodian fin. At right, reconstruction 
of the pectoral girdle of an acanthodian. Cor., coracoid; Sc., scapula; Derm. 
PL, dermal plate connecting pectoral spines. Both from D. M. S. Watson. 

filaments projecting outward into a gill chamber covered by an operculum, is 
achieved." In other words, the acanthodians reached a primitive ganoidal condition. 

We have now surveyed briefly the evidence for and against four of the sugges- 
tions which have been made to account for the origin of the vertebrates. We have 
seen that at least two of these theories point toward the echinoderms as the stem 
forms. It may be worth while, therefore, to summarize the characteristics which the 
echinoderms share with the chordates. 

The radiate symmetry and the plated skeleton of the echinoderms give these 
animals an outward appearance so little like that of a typical vertebrate that it may 
seem absurd to suggest that the two groups had a common ancestry. Although many 
of the former are freely motile organisms, the evidence gained from the study of 
their anatomy, embryology, and history indicates that they owe most of their peculiari- 
ties, such as the radial symmetry, the water-vascular system, and the coil of the gut, 
to the fact that their ancestors were sedentary. All of the motile forms had "pelmato- 
zoic" (sessile) forebears, and show this by retaining a fixed stage in their life histories. 
The earliest echinoderm was probably an animal with mouth at one end, anus at the 
other, a straight alimentary canal, and no skeleton. From it may have been derived 
two groups, one of which adopted an active, the other a sessile, mode of existence. 


One progressive tendency inherited by both groups was that toward activity, for 
from the primitive sessile cystoids, crinoids, and blastoids free-moving descendants 
have sprung. 

The great similarity of the larva of Balanoglossus, the simplest living chordate, 
to the larva of the echinoderms has already been mentioned. Another resemblance 
to the vertebrates lies in the fact that the skeletons in both groups are internal. This 
can be seen most plainly perhaps if a young crinoid be examined. The fact that the 
skeleton is internal is there so obvious and the larval plates of the column with their 
axial canal are so like vertebrae that the similarity is striking. There is also consider- 
able resemblance between the method of formation of the plates of the echinoderms 
and the bones of the vertebrates. The skeleton of the echinoderm is deposited in the 
mesenchyme, where there are amoeboid cells whose pseudopodia fuse into a latticelike 
tissue. Where the pseudopodia meet, the protoplasm secretes small calcareous spicules 
which gradually increase in length; adjacent ones meet and fuse, thus building a 
reticulate structure which in time is augmented to form plates. A similar process 
goes on in the mesoderm of vertebrates, except that the material secreted, instead of 
being calcium carbonate, is largely phosphate of lime. Some other organs of the 
echinoderms also connote the vertebrates. Both groups are coelomate; that is, their 
internal organs are contained in a body cavity. The primary nervous system of both 
is derived from the ectoderm. It remains superficial in the crinoids and starfish, but in 
other echinoderms it sinks below the surface, as it does in vertebrates. Even the liga- 
mental tissue which connects adjacent plates is an intercellular substance, secreted in 
the mesenchyme and like the cartilage of the vertebrates. 

It is, of course, not wise to push these similarities too far as proofs of consanguinity 
between the two groups. Nevertheless, they seem to indicate a sufficiently close rela- 
tionship to remove any feeling that it would be impossible to derive the vertebrates 
from the same ancestors as the echinoderms. 

In the light of what is now known about the ostracoderms it is possible for the 
paleontologist to make a few comments on the theories outlined above. One point 
which impresses him is that perhaps the notochord is not so important a structure as 
has been supposed. The condition found in Silurian and early Devonian cephalaspids 
and anaspids suggests that the tripartite brain and the position of the central nervous 
system above the gut were achieved before the notochord was evolved. This rests, 
it is true, upon negative evidence, the lack of any trace of notochords in the early 
ostracoderms. Since the notochord is composed of soft tissue, one would not expect 
it to be preserved. On the other hand, there does not appear to be any mechanical 
necessity for an internal stiffening rod in animals with strong external skeletons to 
which muscles can be attached. 

So far as is now known, the three groups into which the ostracoderms are divided 
show no connecting links; yet all three of them led to animals with notochords. Since 


it is not likely that a notochord would be evolved independently by each group, this 
indicates that all came from a single stock and that fundamentally their potentiali- 
ties were the same. The stem of the three lines must be sought in ancient rocks, 
not younger than Mid-Ordovician, more likely Cambrian, and possibly even pre- 
Cambrian in age. 

vThe anaspids are of interest in connection with the Amphioxus theory. The 
scaled anaspid is somewhat Amphioxus-\\kt, particularly in its prominent expression 
of the segmentation of the muscles. In most respects, however, it is a more highly 
organized animal. Its pectoral and tail fins are more fishlike than those of Amphioxus, 
and the brains of the two are entirely different, Amphioxus having no vertebrate-like 
brain in fact, practically no head or brain at all. It seems improbable that the anas- 
pids should have descended from any animal with the combination of a poor brain 
and a highly developed notochord. More likely Amphioxus is a retrograde descendant 
of an anaspid-like creature. With our present knowledge of the fossils it seems unlikely 
that any scaleless animal like Amphioxus will be found in the ancestral line. Difficult 
as it is, it is easier to connect the anaspids with true fish than to trace a relationship of 
Amphioxus to them. 

The fossils throw little or no light on the annelid theory. Annelid worms are 
known to have been fully evolved as early as Mid-Cambrian times, but no fossils 
have been found which indicate any connection between them and the ostracoderms. 
It may be that the postorbital valley of the cephalaspids and anaspids represents the 
position of a recently closed-off mouth, and it is true that both anaspids and annelids 
are richly segmented, but neither of these characteristics has much weight in showing 



Anaximander says that men were first produced in fishes, and when they were grown 
up and able to help themselves were thrown up, and so lived upon the land. 

Plutarch, Symposiacs, Book vm, Question ix 

More than half of recorded time had passed before animals emerged from the 
water and tried life on land. During millions of years, invertebrates and fish were the 
sole inhabitants of the globe. Generation succeeded generation, the newborn animals 
in each resembling their parents so closely that a contemporary observer, had there 
been one, would have noted no change, no improvement. Yet there were changes, 
even though they became apparent only after the lapse of thousands of years. We see 
the same process in our children. To us they seem the same from day to day, but 
Grandmother exclaims: "How Ruth has grown since last Thanksgiving!" If we 
stop to think of it, we see two reasons for Ruth's change. One is inherent, produced 
by development with increased age; the other is the effect of environment, partly 
physical and partly mental. Climate, food, and associations all have their effects. 

Our attempts to read the records of which fossils are the symbols cannot be 
completely successful, for we cannot even guess the effects of the mental environment. 
The hopes and fears of the Devonian fish are not measurable. Yet they were not 
inanimate clods, not rocks or minerals, but living animals. "An oyster may be crossed 
in love." We know that they had brains, that they had desires and needs and fears 
and associations with other animals, and that these must have had their effects upon 
the creatures themselves. Since he cannot evaluate this factor, the paleontologist is 
apt to ascribe changes to the physical environment alone. We have an example of 
this in the theories which have been proposed to account for the emergence of the 
fish from the water, and we shall see it repeatedly in further discussions of terrestrial 
animals, for change progress, if it may be so called was much more rapid in the 
years which followed the Silurian than in those which preceded it. The fundamental 
cause of this change was the transition from life in the water to that on land. The 
exodus from the aquatic environment and acquisition of organs permitting life on 
land were epochal events in the history of the vertebrates. From the standpoint of 
the theories of evolution, we are interested to know whether this step was taken in 
response to an "inward propelling force," an urge to essay land life, or whether it was 
motivated by external circumstances. 


In the passage from life in water to life on land two systems of organs must have 
been particularly affected, namely, those of locomotion and respiration. Fish breathe 
by means of gills, paired vascular pouches on the sides of the head, in which the 
blood is separated by a very thin membrane from the oxygen dissolved in the water. 
Adult amphibians, on the other hand, oxygenate their blood by means of air taken 
into, the lungs. 

Fish swim partly by use of fins on the tail, partly by sinuous movements of the 
body, but in addition to the tail fin other similar outgrowths are present which serve 
to assist in balancing, guiding, and, to a lesser degree, in paddling. Of especial im- 
portance in the present discussion are the paired fins, for it is believed that the limbs 
of the terrestrial vertebrates were derived from them. There are several kinds of 
modern fish in which the paired fins are known to be used more or less as legs. Thus 
some are known to creep about on the bottom in shallow water, and even to jump 
above the surface by the use of the fingerlike rays of the pectoral fins, and the "climbing 
perch" of Ceylon holds its place on the roots of trees by use of the same organs. No 
one of these animals, however, shows any structures which even suggest the typical 
limb of the land vertebrate. 

A comparison of the fin of any modern fish with the internal supports of the leg 
of a terrestrial animal will show that it is impossible to derive one from the other. 
The upper arm of a tetrapod (four-legged animal) has a single bone, the humerus; 
the lower arm contains two, the radius and ulna; the wrist has two or three rows of 
short bones or cartilages, the carpals; the palm a row of long ones known as the meta- 
carpals; and in the fingers are several phalanges. The leg shows the same arrange- 
ment, although the bones are given different names. The upper is the femur; the 
lower ones are the tibia and fibula; the ankle contains the tarsals, beyond which are 
the elongate metatarsals and the phalanges of the toes. Since even the most ancient 
four-legged creatures, the amphibians, have this typical arrangement of the bones, 
they are of no help in furnishing links with the aquatic ancestor. It is necessary, 
therefore, to approach the subject from the other side, that of the fish. 

In Devonian times there were two kinds of fish whose fins had a segmented 
internal axial support: the "lobe fins" or crossopterygian ganoids, and the Dipnoi or 
lungfish. With the possible exception of the pleuracanthid sharks, these are the 
only fish ever in existence which had a fin that by any modification could be trans- 
formed into a leg of the tetrapod. At one time it was thought that the lungfish might 
have been the ancestors of the amphibians, but they have been ruled out by their 
specialized, fan-shaped dental plates. There is no possible way in which such teeth 
could have been transformed into the sharp conical ones of the amphibians. Even 
the oldest known Devonian dipnoans had already lost the teeth of the marginal 
bones of the jaws. Once lost, they could not be regained. 

This leaves only the lobe fins as a possibility, for the sharks mentioned in the 


preceding paragraph appeared on the scene too late to be considered. For the past 
twenty years attention has therefore been focused on the crossopterygians. Two genera 
are especially well represented by specimens, the Osteolepis of the Mid-Devonian 
of Scotland and Eusthenopteron (Fig. 53), a common form in the Upper Devonian 
sandstones at Scaumenac in Quebec. A related fish from the Upper Devonian at 
Blossburgh, Pennsylvania, named Sauripterus, has also proved important in this con- 

FIG. 53. The fringed-finned ganoid, Eusthenopteron. From W. L. Bryant. 

FIG. 54. At left, the fundamental elements in the arm and shoulder girdle of Eusthenopteron. 
From W. L. Bryant. At right, above, Broom's diagram of the pre-Sauripterus fin. Below, drawing 
of the fundamental cartilages of the arm of Sauripterus, also by Broom, S. Cl., superacleithrum; 
Cl., cleithrum; Cv., clavicle; Co., Sc., scapula with coracoid; H., humerus; U., ulna; R., radius. 

nection, for the type-specimen presents an unusually well-preserved pectoral fin and 
shoulder girdle. The fins of these three fish are similar in that each has at the 
proximal end a single cartilage corresponding to the humerus of the arm. Articulating 
with its distal end are two cartilages having the position of the radius and ulna. Be- 
yond them are several more, two or three rows, which may be compared with the 
wrist bones (carpals) and palmars (metacarpals) of terrestrial animals. These in 
turn support the rays of the fin. Although the cartilages are not much like the bones 
of the amphibian arm, their arrangement is fundamentally the same. Furthermore, 
the upper one, the humerus, is articulated at its proximal end with another which 


corresponds to the shoulder blade or scapula, an important element of the pectoral 
girdle of all tetrapods. 

So much for the possibilities in the way of limbs. How did the air-breathing 
habit come about? Here paleontology can help less, but the rocks are not entirely 

JEveryone has probably seen goldfish come to the surface to swallow air. Since 
they have no means of aerating their blood effectively by direct contact with the air, 
this is not a method of breathing. Fortunately, however, lungfish are still living in 
Africa, South America, and Australia, and from their habits it is possible to infer 
a reason for the origin of a lunglike organ. All fish other than sharks and skates, 
flounders, and some teleosts which inhabit swiftly running streams have a saclike out- 
growth of the alimentary canal which is known as an air or swim bladder. Its function 
in most fish seems to be to counteract the effects of the various pressures under which 
the animals live at varying depths. In the adults of most modern teleosts it has no 
connection with the oesophagus, but in the ganoids and lungfish it opens directly 
into the gut. Since there are only three species of modern lungfish, living in widely 
separated regions, they may be briefly described. 

Neoceratodus, the Australian lungfish, is a sluggish inhabitant of streams. Be- 
sides breathing by gills, it uses its air bladder as a lung, for at regular intervals it 
comes to the surface, expels foul air, and takes in a fresh supply. During the dry 
season the rivers in which it lives cease to flow, and the fish are confined to pools 
which become foul and deficient in oxygen. At such times Neoceratodus finds the 
lung of great use, for it enables the fish to survive till the next wet season. A very 
closely allied lungfish, placed by some in the same genus as the modern form, has 
been found in the Triassic of England, Germany, India, and South Africa, and in the 
Jurassic and Cretaceous of England, Colorado, and Wyoming. 

The air bladders of the other modern lungfish, Protopterus (African) and Lepido- 
siren (South American), are still more lung-like, since they are double organs. Pro- 
topterus, an inhabitant of marshes in the vicinity of rivers, is large and eellike, some 
specimens reaching six feet in length. Instead of trying to follow the receding water 
into the rivers during the dry season, it burrows into the mud, where it forms a hard- 
ened capsule about itself by means of mucus secreted by the glands of the skin. This 
capsule has an aperture, the margins of which are pulled inward to form a tube which 
ends within the mouth. The fish lives in this condition for nearly half the year, 
breathing through the tube by means of the lungs and living upon the fat which 
it stored up during the feeding season. When thus dormant the animals may be 
dug out and shipped safely to aquaria in any part of the world, to be resuscitated by 
soaking in tepid water. Lcpidosircn has much the same habits. 

The habits of these fish and their use of the air bladder for a lung strongly suggest 
the method of origin of the air-breathing habit among the Amphibia. As has already 


been pointed out, there is a Neoceratodus-like fish in the Triassic rocks. When one 
traces the group further back, it appears that similar ones were present in the Devonian, 
and that the Devonian lungfish were so closely allied to the lobe-finned ganoids that 
it is probable that they were descended from the same ancestor. Curiously, it is known 
that at least one Mesozoic representative of the crossopterygians had a swim bladder. 
Specimens of the Jurassic Undina have been found so preserved that the outline of 
this organ can be seen. By some freak of nature the bladder happened to have its 
wall partially calcified, to the detriment of the individuals cursed with this pathologi- 
cal condition but much to the satisfaction of the paleontologist. 

No calcified air bladders have been found among the remains of the Devonian 
representatives of the lobe fins, but Professor D. M. S. Watson has recently discovered 
evidence that Osteolepis, a Mid-Devonian representative of the group, could breathe 
air. He found that the anterior region of its skull is occupied by a mass of cancellous 
bone in which are a pair of small, nearly spherical cavities which must have been 
occupied by the olfactory organs. Passages extend forward from them to external 
narial openings, and, what is more important, posterior canals pass downward through 
the roof of the mouth, showing that the animal could swallow air just as an amphibian 
does. Evidently it must have depended in part at least upon a lung for the oxygenation 
of its blood. We are justified, therefore, in postulating the existence in early or Mid- 
Devonian times of a ganoid with fins somewhat like those of Sauripterus and lungs 
like those of the dipnoans. This was material from which the Amphibia might 
evolve. What gave the impetus which set the machinery in motion? 

For this it is necessary to look at the physical side of the geological record. Since 
the oldest so-called fish are indeterminate fragments of ostracoderms, the Ordovician 
record is useless, and the Silurian not much better, for remains of fringed-finned 
ganoids of that date have not yet been discovered. In Middle Devonian strata sharks 
are found more commonly in marine than in fresh-water associations, but most of 
the ganoids and lungfish are in fresh-water deposits. It is probable that the hypotheti- 
cal ancestral ganoid lived in rivers and lakes as yet undiscovered. 

It will be noted that modern sharks lack the swim bladder, and show no evidence 
of ever having had one. Is it not possible that this structure arose only in such Lower 
Devonian fish as occupied fresh waters, whereas the sharks are descendants of marine 
animals which did not undergo the particular stresses which led to the formation of 
this organ? 

The Devonian was a time of uplift and mountain-making in northern Europe 
and northeastern America. As the mountains rose, they may have cut off precipi- 
tation on the landward side, so that there ensued a time during which the rainfall 
on their northwestern slopes became deficient and seasonal. This would result in 
alternations of wet and dry seasons, with consequent raising and lowering of the 
water level in lakes, marshes, and rivers, and the production of conditions similar to 


those under which lungfish are living at the present time. Professor Joseph Barrell, 
who suggested this theory, remarks: "The seasonal dryness, with the shrinkage and 
fouling of the fresh waters, or even their complete local disappearance, was a feature 
of the Devonian which appears to have increased in intensity to a maximum in the 
epoch of the Upper Old Red Sandstone" (Upper Devonian). Under such conditions 
it i* obvious that then, as now, the forms capable of making most use of air would 
stand the test chance of surviving. This was probably the time for the development 
of lungs, although there is, of course, no inkling of the modus operandi. The long con- 
tinuation of the conditions which operated to start the air-breathing habit, acting upon 
ganoids equipped with lungs and jointed fins, may have caused them to quit the 
water entirely, or, more likely, may have caused the water to quit them for months 
at a time. Their only chance of survival was as air-breathing, terrestrial creatures. 

This is, of course, only a theory, based on a series of coincidences, but it seems 
to have the support of a considerable number of facts. Lobe-finned ganoids, capable 
of swallowing air, existed during the Middle Devonian. Amphibia first appeared 
in the Upper Devonian. The ganoids and lungfish inhabited fresh water, and the 
red color of the rocks in which their remains are found is considered by many as 
evidence of an increasingly arid climate. Add to this the fact that the early amphibia 
show many ganoid-like characteristics in skull and other parts of the skeleton, and 
Barrell's theory appears at least plausible. It has, indeed, been widely accepted by 
paleontologists and geologists. But in recent years there has been an ever-increasing 
suspicion that the Devonian was not a time of arid or even semiarid climates. Red 
sandstones and shales must have been formed from materials derived from lands on 
which the soils were red. At the present day red soils are not produced in regions 
with a semiarid climate, but in moist tropical or subtropical countries where the 
accumulation of humus is prevented by constant bacterial activity. The writer has 
therefore suggested an alternative hypothesis based upon the idea that the red beds 
of the Devonian were accumulated rapidly during a time of warm and moist climate. 

At present much of the rainfall of the northeastern part of North America is 
precipitated from masses of air which derive their water vapor from the Gulf of 
Mexico. During the greater part of the Devonian the interior of the continent was 
covered by a shallow sea, so that air coming from the southwest should have carried 
even more moisture than now. It may therefore be inferred that the rainfall on the 
western side of the Devonian mountains would have been greater, rather than less, 
than at the present time, a supposition borne out by the fact that the oldest known 
forests are buried in Mid-Devonian strata on the western side of the Catskills. 

If the red beds of the Devonian were really accumulated under moist and warm 
conditions, we must change the theory somewhat, although the explanation re- 
mains the same. The important factor in Barrell's scheme was the evanescent nature 
of the aquatic habitat, and such a situs is supplied by any alluvial fan, under any 


climatic condition. There streams are constantly changing their courses, pools and 
swamps are formed in which the waters are renewed by periodic overflow from the 
main channels, and in these any trapped animals are subjected to the same process 
of gradual exhaustion of oxygen as under semiarid conditions. 

There are certain advantages to be gained from the explanation according to 
the conditions now postulated. Where extreme fouling or actual drying takes place, 
as in the homes of the lungfish of the southern hemisphere, the animals hibernate for 
a part of the year; sloth rather than progress is encouraged. Fish inhabiting the con- 
stantly changing waters of a fan would doubtless often be routed from their homes 
by sudden changes in channel, to be left flopping about in shallow water as the 
streams spread over a flat. Those best able to travel by means of fins might win their 
way to deeper pools. A premium was placed on activity; the weak and unlucky were 
weeded out. The instinct for overland migrations to other and fresher waters may 
early have been developed under such precarious conditions of life. Food, too, would 
have been more abundant and the supply more constant under moist and warm 
than under semiarid conditions. Further, with a greater supply of vegetable matter, 
fouling by decay could occur more readily. Last and most important, amphibians 
have been connected with water throughout their entire history. They constantly 
return to it to breed, except for a few specialized groups of the race, and no theory 
of their origin is satisfactory which requires their complete exclusion from it. 

Since it has become obvious that the crossopterygians and not the lungfish were 
the ancestors of the tetrapods, writers have repeatedly pointed out that the pectoral 
fins of the former contain the fundamental elements of the limb of terrestrial animals. 
Although the shoulder blade and arm bones can be identified, they are associated with 
numerous other cartilaginous or horny fin rays. There are so many of these extra 
elements that the method of their elimination has presented an obstacle to the accept- 
ance of any known lobe fish as an ancestor. BarrelPs suggestion that the fins were 
worn down to stumps as the animals became crawlers instead of swimmers has been 
accepted by many, but has seemed fantastic to others. The Lamarckian doctrine that 
use promotes growth is contravened by this explanation. 

Professor Robert Broom has lately brought forward an idea which appears to be 
a logical solution of the difficulty. He believes that, as the fish came more and more 
into contact with the bottom, the anteroventral part of the fin came to be used in 
crawling, whereas the distal portion retained its normal function (Fig. 54, upper 
right). Sauripterus, according to him, had become specialized for life on the bottom, 
using the ventral part of the pectoral fin for scooping out troughs for itself in the sand 
or mud. Such use is indicated by the flattening and the anterior expansion of the 
humerus and radius. This removes Sauripterus from the ancestral line but serves 
to emphasize the bottom-living habits of the group to which it belongs. Eusthenop- 
teron shows less of this specialization than Sauripterus, having a fin only slightly differ- 


ent from the hypothetical one shown in Broom's drawing. An animal with such a fin, 
forced to adopt a crawling method of existence, would in the course of time lose the 
distal (swimming) portion because of disuse. At the same time, the proximal crawl- 
ing part would continue to increase in size and strength. This theory accounts for 
the loss of the superfluous elements in the same way that the absence of lateral toes 
in^many mammals is explained. 

Barrell surveyed the field for other possible causes of the emergence of the verte- 
brates from the water and suggested three. These are: first, enemies in the water; 
second, food on the lands; and third, the lure of atmospheric oxygen. Enemies of 
the crossopterygians were practically nonexistent. They were the largest, strongest, 
most formidable carnivores of their day. It is true that lack of food or overpopulation 
of the waters may have led them to prey upon each other, but in such a case only the 
weaker members of the community, not the most fit, would have been driven from 
the water. It is not probable that such sudden catastrophic action would lead to 
evolutionary changes. It is more reasonable to suppose that the modifications which 
prepared the fish for life on land occupied a long period. Food on the land could 
hardly have attracted fish unless they were aware of its existence, and it is not possible 
to credit them with reasoning power sufficient to make them conscious of anything 
outside their immediate environment. If the amphibians and their forebears had been 
vegetarians, it is conceivable that feeding on such terrestrial plants as hung in the 
water would have led them to try to climb the banks, but carnivores were exposed to 
no such temptations. Furthermore, so far as is known, there was no food for them 
on the land at that time. Even after they became semi terrestrial they must have re- 
turned to the water to feed. Finally, the lure of atmospheric oxygen cannot be classed 
as an active, driving force. Oxygen was plentiful in the air even in pre-Cambrian 
times, but not till near the close of the Silurian did any animals or plants come suffi- 
ciently into contact with it to develop any organs through which it could be utilized. 
It must have been necessity, not desire, which effected the change. 

Since none of these causes, escape from enemies, desire for the food of the land, or 
lure of atmospheric oxygen, seems to have been sufficiently powerful to lead to ter- 
restrial life, one is forced to return to the theory that it was brought about by changes 
in the environment. In this connection it should be pointed out that scorpions, insects, 
diplopods, gastropods, and plants, all of aquatic ancestry, adopted the air-breathing 
habit at about the same time that the fish did. The environmental changes that affected 
one of the groups may have influenced all, driving them to the air-breathing habit, 
though the result was accomplished variously. 

Whatever we think of the theories outlined above, it is evident that amphibians 
are more closely related to crossopterygians than to any other fish that have ever 
existed. Although no lobe fin yet found is exactly the sort of dnimal which could be 
considered the direct ancestor of the tetrapods, it is probable that some early short- 



finned race not very different from Osteolepis or Eusthenopteron furnished the pro- 
genitor. This is indicated by the fundamental characteristics common to lobe fins 
and amphibia, unshared by the other fish. 

First, and perhaps most important, the Amphibia, and through them all other 
tetrapods, including man, appear to have inherited directly from the crossopterygian 
the fundamental arrangement of the skull bones and the pineal eye. If one examines 
the skull of an amphibian, a reptile, or a mammal, he sees that it is made up of many 
bones united more or less firmly along lines which are known as sutures. The bones 
have an orderly arrangement, basically the same in the three groups mentioned, 
although the skull of a primitive amphibian has many more parts than that of a 
mammal. The bones are symmetrically placed on opposite sides of the median line 

FIG. 55. Diagram to show the similarity of the bones of the top of the 
skull of (A) a crossopterygian, and (B) Ichthyostegopsis, one of the oldest 
known amphibians: n, nasal; fr, frontal; pa y parietal; dso^ dermosupraoccipital; 
first pf, prefrontal; second /?/, postf rental; po, post-orbital; //, inter temporal; 
/, tabulare; /, lacrimal; /, jugal; sq y squamosal. Modified after T. S. Westoll. 

and are easily subdivided into three regions: the dorsomedian, the marginal, and the 
lateral or cheek areas, the last including all between the dorsomedian and the mar- 
ginal series. On the median area (Fig. 55; see also Fig. 64) there are three or four 
pairs of important elements, the nasals, extending backward from the nostrils, then 
the frontals, and back of them, the parietals. Behind the latter are the dermosupra- 
Dccipitals, called by some the postparietals. In the marginal series there are pre- 
maxillaries in front of the nasals; behind each is a long maxillary; these two pairs bear 
teeth. Behind each maxillary is a quadrato jugal, and beneath the hinder corner in 
amphibia and reptiles, but not in mammals, the quadrate, the element with which 
the lower jaw articulates. The number of bones in the cheek series is highly variable 
in both amphibia and reptiles, so only a few of them need be mentioned. Most im- 
portant are the jugal, below the eye, and the squamosal, above the quadrate and the 
quadra tojugal. There may be others between the squamosal and the parietal. In 
front of each eye is a lacrimal and, in most cases, a prefrontal; behind it, a postorbital 
and a postfrontal. 


The similarities between amphibia and crossopterygians are marked. Though 
different skull bones appear in different sorts of crossopterygians the fundamental 
arrangement is that just outlined. Since they are the only fish which have it, it 
indicates that they are directly related to the amphibians. The palatal part of the 
skull, likewise, has the same arrangement of bones in both, as is shown by the ac- 
cojnpanying diagrams. The lower jaws are morphologically identical in crossop- 
terygians and early amphibia, but both possess so many elements that it is not profitable 
to go into a detailed account of their structure. 

The distribution of the teeth is alike in lobe fins and amphibia. It appears further 

FIG. 56. At left, palatal view of Eusthenopteron, for comparison with that 
of a Mid-Carboniferous amphibian at right. P. M., premaxilla; Pal, palatine; 
E. Pt., ectopterygoid; MX., maxilla; Qu., quadrate. Figure at left, from 
W. L. Bryant; at right, after D. M. S. Watson. 

that their emplacement is the same; they are of what is called the acrodont type, 
fixed on the bones without sockets. In replacement new teeth are formed independ- 
ently alongside the old and gradually oust the latter as their bases are resorbed. They 
are simple, conical, and structurally alike, for in both enamel is infolded into the 
dentine, producing what is known as the labyrinthodont type of tooth. 

Another interesting feature, which has only recently been made known by Wat- 
son's work, is that one of the oldest amphibians whose skeleton is known, the Mid- 
Carboniferous Eogyrinus, has a fishlike shoulder girdle attached to the head, like 
that of the crossopterygian. This pectoral girdle consists of the primary support, the 
scapula, and four dermal elements. The latter are external in Osteolepis and have the 
histological structure of scales. In Eogyrinus they appear to have been drawn within 
the skin, but the arrangement is the same. 


Finally, in the most ancient amphibians the dermal bones are not attached to 
the brain case; they are superficial in location, and the body is covered with bony 
ganoid scales, instead of being naked as in modern forms. These many similarities 
are obviously not to be ignored. 

There was really remarkably little change in the anterior part of the body in the 
transformation of a fish into a terrestrial animal. The ornamented surfaces show that 
the cranial bones of the amphibians were practically external, just as in the fish. The 
scales beneath the jaw (gulars) were lost, as were also the large plates protecting 
the neck. The loss of the latter left a notch in the back of the skull over each ear, the 
so-called otic notch. This was probably covered by a tympanic membrane, the new 
arrangement permitting a great improvement in the power of hearing. 

All the evidence points to the lobe-finned, crossopterygian ganoids as the ancestors 
of the amphibians. The geological information at present available indicates that the 
transformation took place in Lower or Mid Devonian times on the alluvial fans west 
or north of mountains which then extended from Pennsylvania across New England, 
Quebec, Greenland, Scotland, and Scandinavia. It is likely that it did not happen sud- 
denly but that there was a gradual evolution. It was caused by a force that influenced 
not one individual fish, but many. Like animals were subjected to like forces in 
similar environments over a great area. Nevertheless, conditions are never exactly 
identical over such an extensive range, and it is probable that the crossopterygians 
of some particular region were more successful in withstanding adverse conditions 
than those in others. Whether this center of origin was in Europe or America is as 
yet unknown; amphibians existed in both regions during early Carboniferous times 
and even earlier in an intermediate area, now Greenland. 

The lobe fins were not the only chordates which experienced this period of stress 
during the elevation of the Devonian mountains. The cephalaspids and Antiarcha 
throve at first in the fresh waters but became extinct at the end of the Devonian. The 
chondrostean (cartilaginous-skeletoned) ganoids arose during this time, learned the 
trick of tiding over particularly adverse conditions by means of the air bladder, but 
failed to gain the land because their fins were not of such a type as to permit them 
to do so. They remained rare and unimportant animals until after the Devonian, 
but were ancestors of the bony ganoids of the Triassic and Jurassic. The latter in turn 
gave rise during the Jurassic to the true teleosts. The sharks, as a group, stayed in the 
sea. Those which essayed the evanescent streams perished; for some reason they 
could not develop air bladders. Some early arthrodires seemed well adapted to the 
fresh-water habitat, but most of them remained in the sea, where they evolved into 
the mighty Dinichthys and Titanichthys. The little sharklike ganoids, the acantho- 
dians, survived till the end of the Paleozoic, but their simple fins were not such 
as could be converted into legs useful in terrestrial life. The lungfish, too, have sur- 
vived, even to the present day, but theirs was evidently a passive resistance; theirs 


was not to do or die, but entrench themselves till better times came. Many a modern 
family has survived, with little honor, in the same way. 

As the paleontologist sees it, it was fortunate that during times of adverse con- 
ditions there was a family of fish, not nice, gentlemanly creatures, but fierce, aggres- 
sive, self-assertive carnivores, which had the strength and vitality to make a struggle 
for , existence. Forced by the conditions under which they lived to swallow air, they 
eventually developed an outgrowth from the alimentary canal which acted as a lung. 
Periodically deprived of water, they used every effort to return to it. The weaklings 
of each generation died. The strong survived, ever gaining strength, increasing their 
lung capacity and perfecting their limbs. He that hath can get more. Eventually 
they conquered their environment. Tetrapods came into being. 


At table I had a very good discourse with Mr. Ashmole, wherein he did assure me 
that frogs and many insects do often fall from the sky, ready formed. 

Pepys, Diary, May 23, 1661 

Amphibians were born at a time when life was hard. Cast up from the waters, 
they still clung to them, returning always to their ancestral home to breed. The 
widespread marshes and swamps of the Carboniferous afforded them an ideal habi- 
tat, and they increased rapidly in numbers, variety, and size, becoming for a brief 
period the most important inhabitants of the earth. In the Permian they were still 
more abundant, but the extraordinary rise of reptiles had already reduced them to a 
secondary position. The Triassic witnessed a decline in variety, although it was the 
period during which they attained maximum size. From the Triassic onward they 
have been relatively lowly and unimportant creatures, showing more or less recession 
from the high position attained by their ancestors. 

Frogs and toads are the most familiar of modern amphibians; it is their life 
cycle which has given the name to the class, for their existence is a double life, although 
not after the manner of Dr. Jekyll and Mr. Hyde. As everyone knows, the eggs are 
laid, and the young born, in water. The larva has a rounded head and a fishlike body, 
ending in a long tail. This tadpole possesses external gills, covered later in life by 
flaps or opercula, as internal gills are developed in the three clefts which open between 
the five branchial arches on either side. At the age of about two months the animal 
reaches a lungfish stage through the formation of a pair of lungs, so that for a time 
it can breathe either in air or in water. Simultaneously with the lungs, the hind legs 
appear, then the fore ones. The gills and tail are gradually resorbed, the gill clefts 
close, and the animal reaches its adult form. Certain other modern amphibians, such 
as the mud puppies and hellbenders, fail to complete the metamorphosis, and retain 
the gilled and tailed condition throughout life. They are not considered primitive, 
but have apparently reverted to ancestral habits and habitat by a simple process of 
arrested development. The frogs and toads represent the order Anura, so called be- 
cause of the lack of a tail. The aquatic forms just mentioned and the terrestrial 
salamanders are known as the tailed amphibia, technically the Caudata or Urodela. 
Still another type of amphibian now living is represented by the caecilians, small, 
limbless, snakelike inhabitants of tropical countries. 

With the exception of the last group, modern amphibians possess scaleless, rather 


slimy bodies and relatively large heads. The skeleton exhibits distinctive features. 
When viewed from above, the skull shows large openings, the result of the failure 
of certain elements to ossify. The short neck has only a single vertebra, which articu- 
lates with paired processes (occipital condyles) on the back of the skull. The ribs are 
short, failing to reach the breastbone, and the pelvis is connected with only a single 
vertebra of the spinal column. Because of the short neck, the shoulder girdle is close 
to the skull. The supports of the fore limbs are relatively complicated, with more 
elements than are found in the same girdle of the more familiar mammals. The 
pelvic girdle, on the other hand, is simple, but unusual in that the anterior (pubic) 
processes of many forms remain cartilaginous. Only two or three exceptions are 
known to the rule that amphibians of all ages have four toes on the front feet and 
five on the hind, provided, of course, that feet are present at all. Aquatic forms have 
cartilages in place of bones in the wrists and ankles, whereas in the fully terrestrial 
species these elements are ossified. Since cartilage is not readily preserved, many 
fossils have a gap between the bones of the fingers and toes and those of the lower 
arm or leg. 

The oldest American amphibians are represented by tracks found many years 
ago in the Mississippian (Mauch Chunk) red sandstone near Pottsville, Pennsylvania. 
A specimen collected near Warren, Pennsylvania, from the Upper Devonian, has 
been described as an amphibian footprint, but it is a single imperfect impression of 
doubtful value. The tracks found near Pottsville show clearly that a five-toed animal 
passed that way while the sand was moist and prepared to take a good impression. 
Many similar tracks occur in strata of late Paleozoic and Triassic age in North 
America, England, and Germany. That they were made by amphibia is shown by 
the impressions of five toes on the hind tracks and four on the front, a broad, hand- 
like outline, and the absence of angular terminals such as are made by claws. 

One of the most important of the many valuable discoveries made by Lauge 
Koch during his recent brilliantly successful expeditions to Greenland was that of 
Devonian amphibians. The material consists of incomplete skulls and other parts 
of skeletons, from which animals named Ichthyostega and Ichthyostegopsis (Fig. 
55 B) have been reconstructed. The names are supposed to indicate that the animals 
were fishlike stegocephalians, that is, intermediate between the fish and the ancient 
amphibians. Unfortunately, they are typical four-footed creatures only slightly more 
fishlike than some of the later members of their group. Primitive conditions may be 
seen in the presence of a few bones not present in later amphibians, among them an 
unpaired rostral element at the front of the skull. The curious position of the external 
narial openings, submarginal, and almost ventral, is superficially like that of the 
crossopterygian fish, but the opening is bounded by different bones. When more 
fully known, these little creatures may prove to have characteristics which justify 
their names. 


Though modern amphibians, except the caecilians, have a naked skin, the late 
Paleozoic members of the class show more or less protective armament. Because their 
heads were covered with bony plates, they are called Stegocephalia, or roof-headed. 
Strangely enough, the armor was thicker on the lower than on the upper surface of 
the body, for some which had naked skin above were covered with scales beneath. 
They varied in length from two inches to eight or nine feet. Most of them had two 
pairs of limbs, but a few were long and snakelike, the legs poorly developed or en- 

FIG. 57. At left, restoration of the primitive embolomerous amphibian, 
Diplovertebron. Two-thirds natural size. From D. M. S. Watson. At right, 
above, a restoration of Branchiosaurus, with (I, II, III) external gills. CAR., 
carpus, and TAR, tarsus, are blank spaces because of failure of the original 
cartilages to be preserved. A little larger than natural size. From Bui man 
and Whittard. At right, below, restoration of three rudimentary branchio- 
saurian vertebrae above the notochord. N, notochord, NC, neural canal, NS, 
neural spine, R, rib. From W. F. Whittard. 

tirely absent. Although the teeth of these early Amphibia were simple cones, many 
have flutings at the base that indicate that the enamel is infolded, as in the lobe-finned 
ganoids. This type of tooth, because of the convoluted structure shown in transverse 
sections, is known as labyrinthodont. Such stegocephalians as possess it are united 
in a large group called the Labyrinthodonta. 

Practically all the labyrinthodonts were large. The head was long, wide, and flat, 
the orbits for the eyes and the pineal opening large. The body was broad, the limbs 
short, with the humerus and femur approximately at right angles to the body. With 
one exception, which has five on both, there were four fingers on the front feet and 
five toes on the hind. These animals appear to have been rather numerous in the 


British Isles during the Upper Carboniferous; a few representatives have been found 
in the American Pennsylvanian. Their remains are relatively common in the Permian 
of Texas and New Mexico, but their greatest development occurred during the Tri- 
assic, when they spread to Germany, Russia, India, South Africa, and Australia. Re- 
cently many of their bones have been found in Triassic strata in New Mexico. 

v The oldest labyrinthodonts, those found in the Lower Carboniferous of Scotland 
and northern England, have been described by Watson. They have a small head, a 
single occipital condyle, four or five cervical vertebrae, a long cylindrical trunk, 
double-ringed (technically embolomerous) vertebrae (Fig. 59 A-C), double-headed 
ribs, and probably, a long compressed tail. The dorsal surface of the head is covered 
with dermal bones, loosely articulated with the case about the brain. The palate is 

FIG. 58. Diplovertebron, one of the few five-fingered amphibians. Restora- 
tion, about one-eighth natural size, by L. I. Price. A different species from 
that shown in Figure 57. 

entirely bony. The shoulder girdle is fishlike, being attached to the skull. Several ani- 
mals of this general type (Embolomeri) are now known, but unfortunately from far 
too few specimens and from too restricted an area to allow as full a knowledge of them 
as could be desired. The structure of their limbs appears to indicate that they were 
terrestrial during their adult lives. 

They seem to have given rise to another group which existed from Mid-Carbon- 
iferous through Permian times, a few persisting into the Triassic. These differ from 
the Embolomeri in having rachitomous vertebrae (Fig. 59 D), each consisting of 
four pieces. The prongs of the neural spines saddled the upper side of the noto- 
chord; a half ring below (hypocentrum or intercentrum) bore the ventral chevrons; 
and the remaining parts consisted of two lateral pieces, the pleurocentra. A good 
example of this type (group Rachitomi or Temnospondyli) is the well-known 
Eryops (Fig. 64 A), several specimens of which have been obtained from the Permian 
of Texas. It is the largest of American Paleozoic amphibians, reaching a length of 
five feet, or if, as is possible, it had a long tail, even more. It has a large head, a heavy 
body, and short massive limbs. The bones of the pelvis are ossified, as in a terrestrial 


animal, but this may be an inheritance from more active ancestors. The position of the 
eyes on the top of the head indicates that it may have awaited its prey partly sub- 
merged in the water like a modern alligator or crocodile. Cacops is a similar animal. 
Closely allied to the amphibia with the four-piece vertebrae is a group with a 
backbone more like that of modern fish. Labyrinthodonts (group Stereospondyli) 
of this sort are characteristic of the Triassic. They were the largest of all, some having 

FIG. 59. At left, dorsal and ventral views of the skull of Pelion, a broad- 
headed branchiosaur, rather common in the American Carboniferous. From 
A. S. Romer. At right, A, B, typical embolomerous vertebrae with neural 
spine saddled on two centra (/', intercentrum, />, pleurocentrum). C, the same 
type with haemal arch attached to anterior centrum. D, rhachitomous caudal 
vertebrae, with haemal arches attached to the intercentra (, neural arch). 
E, a primitive reptile, with small intercentra. F, lateral and dorsal views of 
an intercentrum of a primitive synapsid reptile. G, the same element in a 
rhachitomous amphibian. From S. W. Williston, The Osteology of the 

heads as much as four feet long. Since the lower jaws articulated with backward- 
projecting quadrates, the gape was tremendous. Like modern crocodiles, they may 
have opened the mouth by lifting the upper part of the head, while the jaw rested 
on the mud. They were carnivores, with numerous sharp, conical teeth, two at the 
front of the lower jaws being so long that in some species they passed through holes 
in front of the nostrils of the upper jaws, actually protruding from the top of the 
skull when the mouth was closed. A veritable crocodile among the Amphibia! All of 
the late Triassic members of this group appear to have been aquatic or semiaquatic, 
for they have short feeble limbs. 

The large labyrinthodonts were only one branch of the roof-headed amphibians. 
Other branches, among them the branchiosaurs (Phyllospondyli), microsaurs, and 
Adelospondyli, were represented by small creatures, seemingly inconspicuous and 
feeble, yet in the sequel more important than the labyrinthodonts, for they appear 
rn have furnished the ancestors of the modern amnhihians. 


The branchiosaurs are small, primitive stegocephalians which appeared as early 
as the Devonian. The osseous part of the backbone is simple, composed of what are 
known technically as the phyllospondylous type of vertebrae. The notochord is con- 
tinuous and not enclosed in bone, only the neural arches being ossified (Fig. 57, 
right, below). Such vertebrae are even more primitive than the double-ringed ones 
of ttye embolomerous type. The teeth are conical; some species show simple infoldings 
of the enamel, suggesting the convolutions of the labyrinthodonts. All members of 
this group have large, broad, flat heads, short bodies and tails, and feeble limbs, 
which indicate that they were chiefly aquatic (Fig. 57, right, above). The ventral 
surface of the body was covered with thin hemicycloid scales; the upper side was 
naked. The large short head, straight ribs, short tail, large openings in the palate, 
cartilaginous wrists, ankles, and pectoral girdle parallel the structure of modern 

FIG. 60. A restoration of Cacops, one of the large Permian Rhachitomi of 
Texas. More terrestrial than most of its relatives, it had the rudiments of a 
dorsal armor, and rather efficient legs. Original drawing by L. I. Price. 

frogs. Specimens of branchiosaurs have been found at Linton, Ohio (Fig. 59, left), 
and Mazon Creek, Illinois, in Mid-Carboniferous strata, but they are best known 
from the abundant material collected from the Permian near Dresden, Germany. 
More than a thousand specimens of Branchiosaurus from the latter locality were 
studied by Professor H. Credner. He found that many immature individuals showed 
external gills (Fig. 57, right, above; Fig. 61), supported by projections from the 
gill arches, which seem to have been at least partially calcified (Fig. 62). When 
the larvae reached a length of about 100 mm. (four inches), they lost these structures. 
This change was accompanied by a reduction in the length of the tail, which in the 
young had been as long as the body, an expansion of the pelvis, and an increase in 
the ossification of the skull and other parts of the skeleton. One sees in this a begin- 
ning of the metamorphosis which plays so conspicuous a part in the life history of 
a modern frog. 

Somewhat like the branchiosaurs, but probably not derived from them, are the 
microsaurs (Lepospondyli), small animals with "lepospondylous" or hourglass-shaped 


vertebrae. The notochord was continuous, but constricted by the partial ossification 
of the centra of the vertebrae. In general appearance they were much like the 
branchiosaurs, but the body and tail were longer. The hind limbs were longer than 
the fore; the pubes were ossified, an unusual feature in amphibians other than the 
primitive labyrinthodonts; and in many the wrists and ankles had true bones. The 

FIG. 61. Restoration ot a larval branchiosaur, with external gills and a 
long tail. Two-thirds natural size. Original drawing by L. I. Price. 

FIG. 62. A, gill-supports of a larval branchiosaur, with those of a modern 
axolotl, B, for comparison. The elements in black are ossified, the others 
cartilaginous. After H. Credner. 

ventral armor consisted of oval rather than semicircular scales. Most of the group 
must have been fully terrestrial. 

Although the history of this group is at the moment somewhat obscure, it is 
evident that it began rather early in the Carboniferous, for it includes a Mid-Carbon- 
iferous branch known as Aistopoda. The fact that in the microsaurs the fore limbs 
were shorter than the hind suggests a "degeneration" of the appendages. Various 
transitional forms led to limbless, snakelike creatures (Fig. 63), numerous in Car- 


boniferous strata of Ohio and Ireland. These animals had short ribs and elongate, 
hourglass-shaped vertebrae with centra so nearly solid that there is only a small 
perforation for the notochord. Some had as many as sixty, others one hundred seg- 
ments in the backbone. There is some evidence that their bodies were completely 
covered with scales, so they must have looked much like modern caecilians, although 
no^ closely related to them. All were small, from three to ten inches in length, and 
probably aquatic in their habits, for the neural and haemal spines of the posterior 
vertebrae are elongate, as if they supported a tail fin. 

Other possible descendants of the early microsaurs are members of a small group 
known as the Adelospondyli. The vertebrae appear to be a modification of the 
hourglass type, the centra much like those of the Aistopoda. The neural spines are 
not attached to the centra. A well-known member of this group, Lysorophus, an 
animal found in the Permian red beds of Texas, has been much discussed by paleon- 

FIG. 63. A restoration of Dolichosoma, one of the snakelike aistopodan 
amphibia. One-sixth natural size. From Anton Fritsch. 

tologists. Some have thought that it was a reptile; others that it was an amphibian, 
but probably an ancestor of the reptiles. Fortunately, a couple of nodules containing 
skulls and portions of skeletons of this creature were given to Professor W. J. 
Sollas at Oxford. Sollas has his own way of studying fossils, a method which has 
yielded important information in groups as widely separated as graptolites and rep- 
tiles; this is, in essence, simply the use of serial sections. The zoologist cuts sections 
with a microtome; Sollas mounts fossils, still embedded in the original matrix, on a 
frame in such a way that he can grind them on a wheel in planes strictly parallel. 
Each newly revealed surface is drawn or photographed, and thus the entire structure 
is determined. Studying Lysorophus in this way, he demonstrated that it was truly 
an amphibian, with branchial arches like those of urodeles and no pineal opening. 
This seems to indicate that the modern Caudata, but not the reptiles, may be de- 
scendants of the Adelospondyli. Specimens of other small animals belonging to this 
group have been found in the Lower Carboniferous of Scotland and in the Mid- 
Carboniferous and particularly the Permian of both Europe and North America. 
The modern tailed amphibians differ from all the "roof-headed" in lacking 


scales, in having extremely short ribs, and in having lost various parts of the skull. 
Several of the elements regularly fail of ossification, whereas others are bony in 
members of some genera, cartilaginous in others. As a result, the skull appears to 
have a sketchy framework, in marked contrast to that of the stegocephalians. A 
seemingly trivial but important characteristic is the absence of a pineal foramen. The 
limbs are short, and the pectoral girdle is chiefly cartilaginous, as are the pubic 
elements of the pelvic region. The fact that haemal arches are attached to the centra 
of the caudal vertebrae indicates that it is the anterior of the primitive rings which is 

Although most of these creatures are aquatic, they are rare as fossils. A single 
specimen has been found in the Mesozoic, the Cretaceous of Belgium. It is the oldest 
known mud puppy, but it is closely allied to the famous Andrias scheuchzeri of the 
Miocene of Oeningen. The latter is Johann Jacob Scheuchzer's Homo diluvii testis, 
whose story is told to all students of paleontology; the "man" was about three feet 
long and differed but slightly from the giant mud puppy (Megalobatrachus) still 
living in Japan. Several extinct genera of salamanders, differing but little from 
modern ones, have been described from specimens obtained in the Tertiary of France, 
Germany, and Bohemia. The oldest of these terrestrial caudates is of Lower Cretaceous 

The most specialized of all amphibians are the frogs and toads. The Anura have 
an advanced type of vertebra, that known as opisthocoelous because the posterior end 
of the centrum is concave, the anterior convex. Only ten or twelve vertebrae are 
present in the trunk. They bear stout transverse processes, but ribs, except in one 
family, are absent. The caudal vertebrae are coalesced into one short piece, the coccyx. 
The skull shows certain peculiarities; the orbits are large, the parietal and frontal bones 
are fused together, and a platelike bone is interposed between the frontals and nasals, 
The teeth are small, bristlelike, or entirely wanting. The pectoral arch is partially 
ossified, the fore limbs long, with coossified radius and ulna, and ossified wrist bones, 
an unusual feature in the Amphibia. The hind limbs are long, the tibia and fibula 
united, and the ankle ossified. 

The oldest frog is known from specimens found in the Triassic of Madagascar. 
It is so like modern ones, however, that it is obvious that the group originated much 
earlier. There is, as has been mentioned, a suggestion that it was derived from 
the Paleozoic branchiosaurs. A great many specimens have been obtained from 
Tertiary strata of Europe, some of them so remarkably preserved that histological 
studies of their tissues have been made. Phosphatized mummies have been found 
in France, and Miocene deposits near Bonn have produced both adults and tadpoles 
in abundance. 

Absolutely nothing is known of the geological history of the glass snakes, the 


To recapitulate: the amphibians, arising from the lobe-finned ganoids probably 
in Silurian or early Devonian times, inherited from their ancestors a skull consisting 
of numerous bones, the dermal elements of which were attached but loosely to the 
brain case; labyrinthodont teeth; a functional pineal body; a continuous notochord, 
only partially enclosed by bone; a largely external pectoral girdle; the rudiments of 
two pairs of limbs; a long tail; and a covering of scales. 

The general trend of evolution in the group appears to have been toward a 
reduction in the number of bones in the skull, the closing of the pineal opening, the 
constriction and elimination of the notochord through the formation of the centra 
of the vertebrae, the withdrawal of the dermal bones of the skull and of the pectoral 
girdle beneath the skin, the ossification of all the bones of the limbs, or, secondarily, 
their loss, a reduction of the length of the tail, and the loss of the external scales. 
Some of these trends had reached fulfillment early in Carboniferous times, when the 
real record of the amphibians begins. The limb bones were fully ossified; in fact, 
many of the Carboniferous and Permian forms have ossified wrist and ankle bones, 
even ossified pubic bones. Other trends, such as the reduction of the number of 
bones in the skull, perfection of the centra of the vertebrae, loss of the pineal foramen, 
and loss of scales, did not reach their culmination till well into the Mesozoic. 

All of the Paleozoic amphibians were roof-headed stegocephalians. Their truly 
progressive line was that of the labyrinthodonts, which reached their culmination 
shortly before they became extinct in the Triassic. The collateral lines are represented 
by the branchiosaurs, which may lead to modern frogs, and the more terrestrial micro- 
saurs, possibly ultimate ancestors of the salamanders. It remains, however, for paleon- 
tologists to find in late Paleozoic and early Mesozoic strata proof of the ancestry of the 
modern amphibians. 

One of the principal lessons to be learned from the study of the history of this 
group is that degeneration does not necessarily result in degeneracy. The amphibians 
reached the peak of their form during the late Paleozoic, but their most highly spe- 
cialized representatives are the modern frogs and toads. No one could call the latter 
degenerate in any accepted sense of the word. They lead a free, nonparasitic existence, 
and their insect-catching habits make them acceptable members of the modern com- 
munity. Aloysius, our farmstead friend, comes to the outside faucet for his daily 
bath and then establishes himself beneath a grapevine to catch flies. Degenerate? 
Not at all. But he lacks the size, some of the skull bones, teeth, cervical vertebrae, 
ribs, and the tail of his Carboniferous ancestors. Structurally he is degenerate. Prob- 
ably he has neither more nor less brain. If pressed, he can jump a hundred times 
further than any Carboniferous amphibian, but he climbs the back steps in the 
same way as any other four-footed animal. He has a mixture of primitive, "sub- 
primitive," and specialized characteristics. It is the "sub-primitive" characteristics 
which zoologists commonly designate as "degenerate." If they represented reversions 


to an ancestral condition, they might well be called recessions. Unfortunately some 
of the so-called degenerate characteristics are distinctly different from any possessed 
by the ancestor. They represent progress, but progress backward. Perhaps "retro- 
grade" is as good a word as any we have to apply to such lines of evolution. If this 
term be used, the aquatic caudates are simple-retrograde, the salamanders semi- 
retrograde, and the Anura specialized-retrograde. 


These are our blood and bone that climb and crawl 
Up from the mire through the Neanderthal. 

Humbert Wolfe, The Uncelestial City 

Many people have an inherent dislike for reptiles, a prejudice fostered by age- 
long tradition. This antipathy is directed chiefly against snakes, abundant and con- 
spicuous modern representatives of the class. The paleontologist is little interested in 
snakes, for their history is brief; they are a specialized branch of a good old stock, 
their ancestors having been the principal inhabitants of the earth during the middle 
period of its history. From the beginning of the Permian till the end of the Mesozoic, 
reptiles dominated the life of land, sea, and finally air. Our own ancestry reaches 
back to them; so, even though their glory has departed, we must look with a certain 
interest upon a race whose supremacy persisted longer than that of any other single 
group. In retrospect it seems almost as if nature had been experimenting, trying out 
the possibilities of the four-footed creatures she had evolved, seeing what could be 
done by making various modifications of the fundamental structures. But the rep- 
tiles were handicapped by the cold blood their ancestors had brought with them 
from the water. A long period of life on land, eons of adversity, of pursuit of prey, 
of struggle to escape destruction, were necessary for the evolution of that warm blood 
which enabled mammals and birds to displace their sluggish forebears. 

To understand the course of evolution it is, as usual, necessary to delve somewhat 
into details of structure which might seem technicalities important only to the special- 
ist. The general results of the studies of paleontologists since the time of Cuvier 
could be summarized in a few pages, but we prefer to investigate the evidence on 
which the findings are based. 

Reptiles are cold-blooded animals with bodies commonly encased in scales, 
although some are naked. Some have bony plates in the inner layer of the skin, be- 
neath the outer epidermal scales. Incidentally it may be remarked that the reptilian 
scale appears not to be an inheritance from the amphibians. Paleontologists, however, 
are interested chiefly in the skeleton. This in most cases is fully ossified; the vertebrae 
usually have solid centra, for, except in the oldest and simplest reptilian forms the 
notochord does not persist in the adult. Since the skull is variously formed in dif- 
ferent groups, more will be said of its construction on later pages. The teeth are 
simple, conical in the majority but variously modified according to the diet. All 


reptiles, saving a few dinosaurs, have single-rooted teeth. At the back of the skull 
there is a single condyle for articulation with the first of the vertebrae. It is a rounded 
knob beneath the opening through which the spinal cord issues from the brain case. 
The neck of some reptiles is long, that of others short, so that there is great variation 
in the number of cervicals. It will be remembered that most Amphibia have only 
one; most mammals have seven. The shoulder girdle is rather complicated and con- 
sists of more bones than are found in the same structures in placental mammals. 
The pelvic girdle is fully ossified, the upper elements (ilia) connected with at least 
two vertebrae, which are joined together to form a sacrum. Many reptiles have three 
sacrals, a few as many as eight or nine. 

Primitive reptiles have five toes on both front and hind feet, but the number is 
reduced in specialized groups. Oddly enough, one of the most constant reptilian 
characteristics is the number of bones in the digits. Although the term "phalangeal 
formula" seems technical, its meaning is simple; it is merely an easy way of expressing 
the number of bones in fingers and toes. One can readily learn his own formula by 
counting the segments of his fingers. He should remember, however, that in walking 
in the normal mammalian position, the big toe and thumb are inside, the little toe 
and little finger outside. Digit number one, the thumb, has two bones; all the others 
have three each. Hence our phalangeal formula is 2,3,3,3,3, which? incidentally, is 
the typical formula for primitive mammals. If one investigates the bones in the 
fingers and toes of reptiles, he finds that the thumb has two, that there is increase by 
one each to the fourth finger, then a decrease of one segment; therefore the formula for 
the hand is 2,3,4,5,3. That of the foot is 2,3,4,5,4, the little toe having one more bone 
than the little finger. Not all reptiles have exactly this arrangement of bones, but since 
the majority, including the oldest and most simple, do, these are the primitive formulae. 

The oldest reptilian remains are found in rocks of Mid-Carboniferous age. For 
a brief period, many years ago now, the writer, a student of invertebrate fossils, en- 
joyed the enviable reputation of being the discoverer of the oldest known reptiles, 
found in Mid-Carboniferous strata east of Pittsburgh. But glory is transient. As soon 
as their description was published, the vertebrate paleontologists began to rummage 
among their stores and soon discovered that Cope many years earlier had described 
as a reptile a partial skeleton found in a coal mine at Cannelton, Ohio. This coal was 
formed a few thousand years earlier than the Mid-Carboniferous strata which held 
the bones found near Pittsburgh; the specimen from Cannelton, for which Williston 
suggested the name Eosauravus^ is therefore older, and till the present date, 1937, 
remains the oldest reptile known (Fig. 69). The skeleton is a small one without 
skull or fore limbs. Although Cope had correctly identified it, later students con- 
cluded that it was an amphibian. Once the fact was demonstrated that reptiles did 
exist as early as Mid-Carboniferous, restudy showed that the great naturalist was right. 
The only obvious characteristic which shows that it is a reptile is the phalangeal 
formula of the hind foot, 2,3,4,5,4. 


Little is yet known about Carboniferous reptiles, but Permian strata, particularly 
the red beds of Texas, continue to furnish specimens, and rocks of the same age in 
Europe, Africa, and South America have yielded their contributions. The most 
primitive fossils from these beds belong to a group known as the cotylosaurs. The 
present evidence suggests that these primitive reptiles originated in the north, perhaps 
in that ancient Eria which bounded the northern Atlantic. They spread in Carbon- 
iferous and early Permian time to North America and Europe, and during the middle 
and late Permian to South Africa. They disappeared from North America at the 
end of the Paleozoic, but remains of them are found in Lower and Middle Triassic 
rocks in South Africa and Europe. Nowhere did they survive later. 

That some sort of stegocephalian was the ancestor of the reptiles is amply proved 
by the skeletal structure of the cotylosaurs. The skulls are depressed and completely 
roofed with bones; otic notches are present in some of them. Detailed comparison 
reveals that one or another shows every paired bone which was present in the skulls 
of the labyrinthodonts; in general, however, there are fewer bones in the skulls of 
primitive reptiles than in those of amphibians (Fig. 64). The backbone also is 
amphibian-like, for the individual centra are biconcave (amphicoelous), and the fact 
that the centrum is perforated in most shows that the notochord, although much 
constricted, was continuous. 

The outward appearance of the cotylosaurs must have been much like that of 
the contemporary amphibians. Like them, they were semiaquatic, living in swamps 
and marshes, crawling about with their plump, rounded bodies almost dragging on 
the ground, their short legs being inadequate for rapid or energetic movement. 
The skin is practically unknown, but horny scales probably covered the body. A 
few had bony plates serving as a partial protection. The teeth were conical, or modi- 
fied cones, in some forms blunt, adapted for crushing. All appear to have been flesh- 
eaters but not fierce carnivores. It seems, rather, that some of them fed on crusta- 
ceans or insects. The great cockroaches of the day, some four inches long, were 
doubtless appetizing and succulent morsels. Other reptiles were probably fond of 
snails and mussels, their teeth being well adapted for crushing shells. In Texas they 
lived on a lowland facing the sea but got their food on the land and in the fresh 
waters. Most of these animals were small, from six inches to four feet long, although 
a few reached the length of ten feet. 

A few of the cotvlosaurs are so well known, at least among paleontologists, that 
they deserve special description. 

Seymouria (Fig. 64), an animal about two feet long, represented by nearlv com- 
plete specimens from the Permian of Texas, has been widely heralded as the most 
primitive known reptile. It was a short-legged, crawling creature, its widely spaced, 
sharp, conical teeth especially fitted for catching insects. The short neck, massive 
shoulder girdle, and numerous bones of the skull suggest its near relationship to the 


amphibians whose habitat it shared. So amphibian-like is it that one specialist on 
early reptiles, Professor Broom, still insists .that it is merely an amphibian which has 
certain reptilian characteristics; on the other hand, American students believe that 
it is a true reptile. When doctors disagree, the layman is not particularly interested 
in whose opinion is correct but in the reason for the quarrel. From our standpoint 
the significant fact is that here again we find evidence that late Paleozoic reptiles 
were closely allied to amphibians. 

The most important characteristics of Seymouria are those of the backbone. 
The vertebrae suggest the two-ringed, embolomerous type of the early labyrinthodonts, 
but there is a marked difference, for in Seymouria it is the anterior part, the inter - 

FIG. 64. Diagrams to permit comparison of skull bones of (A) Eryops, a 
Permian amphibian, and (H, C) Seymouria, a primitive Permian cotylosaur: 
pm, premaxilla; w, nasal; fr, frontal; pa, parietal; dso, dcrmosupraoccipital; 
/, lacrimal; /?/, prefrontal; pof, postfrontal; //, supra temporal; t, tabulare; 
mx y maxilla; ;', jugal; sq, squamosal; qj, quadrato-jugal; q> quadrate; po, 
postorbital; //, inter temporal; ot, otic notch. Note the unpaired bone on the 
median line in A. A, after R. Broom; B, C, after S. W. Williston, with 

centrum, which is reduced, whereas the posterior ring (pleurocentrum) is complete 
and supports the neural arch (Fig. 59 E). This indicates the workings of a process 
exactly opposite 19 that in the amphibians, where the posterior ring tended to dis- 
appear. All the earliest reptiles appear to have had vestiges of the anterior part of 
the primitive double-ringed vertebra, but in most lineages they disappeared during 
the early or middle Permian, although they have persisted till the present in the 
"living fossil," Sphenodon, and in the geckos among the lizards. Evidence* of their 
former presence is shown by the position of the haemal arches of the tails of many 
reptiles. These appear to alternate with the vertebrae; that is, they are intercentral in 
position, souvenirs of the lost anterior halves of the centra. 

It is unlikely that the discrete pleurocentra of the rhachitomous vertebra would 
grow together again to form a ring; therefore it is not at all likely that the vertebrae 
of the early reptiles could have been derived from a four-piece type. The vertebra 
of a cotylosaur seems to have been formed directly from an embolomerous type by 


the loss of the upper part of the anterior of the two primitive rings. The ancestor 
of the reptiles must therefore be sought among the Embolomeri of the Lower Car- 
boniferous. Only four or five genera are yet known. One had five fingers, but un- 
fortunately not enough phalanges. Further discoveries in the older coal measures of 
Scotland or northern England will probably reveal the form with just the proper 
characteristics. If Diplovertebron (Fig. 57, 58) had a phalangeal formula of 2,3,4,5,3, 
instead of 2,3,3,3,4, lt would doubtless now be heralded as the most important of all 
amphibians, the ancestor of the reptiles. By what narrow margins is glory missed! 
Three little useless bones in the hand! 

Pareiasaurus, represented by many skeletons from the Upper Permian of South 
Africa and northern Russia, is a good example of an Old World cotylosaur. It has a 
massive skeleton, six to eight feet long, short but sturdy legs, strong, clawed toes, and 
a broad heavy skull. This animal, like some other reptiles inhabiting the same regions, 
stood higher than its American cousins. It was one of the first to begin to solve the 

FIG. 65. A, the skull of the Triassic cotylosaur, Elgtnia, compared with 
(B) that of the modern horned toad, Phrynosoma. 

problem of getting the legs beneath rather than beside the body. This was accom- 
plished by a change in the axial direction of the upper arm and leg bones which 
brought them into a vertical rather than a horizontal position. If some of the restora- 
tions of Pareiasaurus may be trusted, this change was a necessity, since otherwise 
the animal could hardly have moved its exceedingly corpulent body. A more funda- 
mental explanation of the change may be seen, however, in the altered positions of 
the various bones of the pectoral and pelvic girdles which resulted in changed direc- 
tions for the muscles. The fusion of the pelvic bones greatly strengthened the support 
for the limbs. Pareiasaurus is supposed to have been a herbivore. 

The last cotylosaur to be mentioned is Elginia (Fig. 65), remains of which are 
found in the Triassic sandstone near Elgin, Scotland. Only the skull is known, 
pieced together from plastic casts made from natural molds in sandstone. Its chief 
interest lies in the fact that its head instead of being smooth, like those of most of 
its relatives, is covered with spinelike protuberances, those along the posterior margin 
being elongated into conspicuous spines. Its skull was in some respects like that of 
the modern horned toad, which is a reptile, not a toad. The late Professor C. E. Beecher 
of Yale, who was especially interested in spinescence, investigated its development 
and history in many groups of animals and plants, with the view of learning its 


significance. His studies led him to the conclusion that in general the production 
of spiny outgrowths is confined to the later periods of the history of any race, occur- 
ring usually only shortly before the disappearance of a particular group. 

Although the cotylosaurs themselves disappeared with the closing of the Mid- 
Triassic, they left numerous descendants, the least modified of which are the turtles, 
discussed briefly in a later chapter on aquatic reptiles. If one wishes to envisage the 
fully roofed skull of the primitive reptile, he should inspect that of a large marine 
turtle. Possibly it is because most turtles, like their cotylosaurian ancestors, are semi- 
aquatic that some have retained a primitive arrangement of the skull bones. All 
other living reptiles show one or two pairs of openings (vacuities) behind the eyes. 
These fenestrae are considered so important that the division of the group into five 
subclasses is based upon their distribution. This grouping was first suggested by 
E. D. Cope, first applied by H. F. Osborn, and fully amplified into a definite scheme 
by S. W. Williston, a paleontologist with a wide knowledge of fossil and recent 
reptiles. In no other group, perhaps, is a knowledge of fossils more important in 
proposing a classification. No living reptile is truly primitive; most are highly spe- 
cialized, and there are many more groups of extinct than of living members of the 

All that tread 

The globe are but a handful to the tribes 

That slumber in its bosom. 

All classifications published before that proposed by Williston are unsatisfactory 
because they fail to recognize that those reptiles are most primitive which are most 
like the stegocephalians from which they were derived. 

The cotylosaurs are obviously the parent stock, for they are the oldest, and but 
slightly different from the contemporaneous amphibians. They show no openings in 
the skull behind the orbits (Fig. 66 A). Like the early amphibians, they are roof- 
headed. How were the reptiles with one or two pairs of temporal openings derived 
from them? 

The mechanics of the explanation are simple. It must not be supposed that be- 
cause the stegocephalians, the cotylosaurs, and the turtles had big, broad heads they 
had big brains occupying the space beneath the roof of the skull. On the contrary, 
their brains were small, enclosed in a narrow, bony box beneath the median portion 
of the parietals and frontals, bones adjoining the mid-line of the skull. The great 
area of the inner surface served merely for the attachment of muscles, particularly 
those working the jaws. The operation of these muscles exerted a considerable stress 
upon the roof bones, pulling them downward and inward. They would naturally 
be most apt to give way along their margins, the sutural union being inadequate. 
It will be remembered from the analysis of the skull bones of the amphibians (Chap- 
ter XI) that three series are readily recognized: a median series, two marginal ones, 


and lateral ones between them. If the bones were pulled apart along the sutures 
between the median and the lateral series, a pair of dorsal openings would be formed: 
if between the laterals and marginals, the vacuities would be at the sides. If they were 
severed along both lines of weakness, both dorsal and lateral fenestrae would appear. 
All these things have happened, and the arrangement of the openings is char- 
acferistic of the subclasses. The reptiles with none, the primitive cotylosaurs and the 
somewhat specialized turtles, form a group known as the Anapsida. Those with a 
single pair of lateral ones, the Synapsida (Fig. 66 B), appear next in sequence. These 

FIG. 66. Diagrams to show lateral views of the skull of (A) an anapsid 
and (B) a synapsid reptile. Lettering as in Fig. 64. After S. W. Williston. 


FIG. 67. Diagrams to show lateral and dorsal views of skulls of (A, B) a 
parapsid and (C, D) a diapsid reptile. Lettering as in Fig. 64. After S. W. 

are all extinct but of great importance, for among them are the ancestors of the 
mammals. Next are the forms with a pair of dorsal openings, divided by Williston 
into two groups, the Parapsida (Fig. 67 A, B), a subclass which includes the ichthyo- 
saurs, and the Synaptosauria, the plesiosaurs and their allies. Last, and largest, is 
the subclass known as the Diapsida (Fig. 67 C, D), with dorsal and lateral temporal 
vacuities which tend to reduce the posterior part of the skull to a sketchy framework. 
Here belong many of the living reptiles, the lizards, the snakes, the primitive Spheno- 
don, the alligators and crocodiles, and their first cousins, the extinct dinosaurs; here 
also are the most specialized of all reptiles, the flying pterosaurs, and early members 
of the subclass are supposed to have been ancestors of the birds. 

Broom has made the ingenious suggestion that the synapsids were derived from 


cotylosaurs with broad heads, and the parapsids from narrow-headed forms, the more 
direct downward pull in the latter resulting in dorsal rather than lateral openings. 
The diapsids, according to him, probably were derived from some parapsid which 
had become secondarily broad-headed. Since the lateral temporal openings seem 
to have appeared before the dorsal, it is possible that the diapsids were derived from 
narrow-headed descendants of the synapsids. This would be more in accordance 
with what is now known of the geological record, for synapsids were common in 
Permian times, whereas the parapsids were chiefly Mesozoic. The remains of the 
oldest diapsid, Youngina, are found in Permian strata in South Africa. 

(Cope). Original, seven and a half feet long. Note the compressed heads, 
as compared with the depressed ones of cotylosaurs. Original drawing by 
L. I. Price. 

Not all paleontologists agree that the temporal fenestrae were formed by muscular 
stresses. Some think the apertures were produced by the resorption of the bones to 
which the muscles were attached. This seems to the uninitiated a bit like the action 
of a man in a tree who saws off the limb on which he is seated, but resorption of bone 
and shell is undoubtedly a much greater factor in change of skeletal form than has 
been generally recognized. R. T. Jackson has for years advocated more study of this 

Most Paleozoic reptiles belong to one or the other of the two primitive sub- 
classes, the Anapsida and the Synapsida. The anapsids, the cotylosaurs, have already 
been discussed. Their companions in the American swamps were the pelycosaurs, 
readily recognizable as synapsids by the presence of lateral temporal openings. 

FIG. 69. The oldest known reptile, Eosaztravus. From Roy L. Moodie. 

FIG. 70. Edaphosattrus, the "telegraph-pole" lizard. A microcephalous, dis- 
harmonic, unadaptable type of pelycosaur. Specimen mounted in the Museum 
of Comparative Zoology, Harvard University, under the direction of A. S. 
Romer. Mount and photograph by George Nelson. 

FIG. 70A. The oldest known egg, possibly that of an Ophwcodon, a Permian 
reptile. Courtesy of Alfred S. Romcr. 


Most of the pelycosaurs were externally similar to the cotylosaurs, their immediate 
forebears. Their skulls were, on the whole, somewhat narrower and higher, com- 
pressed rather than depressed (Fig. 68). Although they were short-legged, they suc- 
ceeded in carrying their bodies off the ground. In these respects they show further 
departure from the form and habits of their amphibian ancestors than the cotylosaurs. 
Probably they were all carnivorous, for most had long, sharp, conical teeth. The 
exceptions are those with a dental apparatus consisting of numerous small, blunt, 
marginal, and palatal cones, adapted for crushing thin-shelled mollusks and crus- 
taceans; only a few have been suspected of being herbivorous, and the case has not 
been proved against them. Most were small, inconspicuous creatures, two or three 
feet long, unimportant members of their community. They received no plaudits in 
their day, nor do they now, as one looks hastily at their skeletons or reconstructions. 
Yet in the fullness of time it was their stock which gave rise to the mammals, and 
their blood flows in our own veins. Man ignores the John Smiths in his ancestry. 

Yet the group is not entirely neglected. Museum visitors pause before skeletons 
of Dimetrodon and Edaphosaurus (Fig. 70), the pelycosaurian freaks, ancestral to 
nothing. Tremendously elongated neural spines on their vertebrae gave these ex- 
traordinary reptiles a bizarre appearance. It is easy to distinguish the genera, for 
Dimetrodon has simple spines and a large, long-toothed skull. He was a predaceous 
carnivore. Edaphosaurus, the "telegraph-pole lizard," was more specialized, for it 
had cross bars on the spines. But the skull is small, and the teeth indicate a diet of 
small invertebrates. The two genera obviously result from two independent lines of 
pelycosaurian evolution. The function of the elongated spines is still unknown. They 
were hardly strong enough to stand alone, so it is inferred that they were inside a 
continuous membrane which formed a dorsal fin. But why should a terrestrial animal 
have such a fin? It was too weak to be protective. Cope long ago suggested that the 
animals were aquatic and used the fin as a sail. The relatively feeble limbs of Eda- 
phosaurus may support this hypothesis, but neither skeleton shows any really aquatic 
adaptations. Lacking other suggestions, paleontologists have fallen back upon the 
orthogenetic idea that, once started, the growth of neural spines continued to a climax 
in which they were overdeveloped and led to the extinction of their owners. This is 
in line with Beecher's ideas about spinescence in general. Neverthless, this explanation 
somehow fails to explain. 

When the writer looks at the restorations of Edaphosaurus^ and realizes how 
microcephalous that snail-eating creature really was, he sometimes regrets that Pro- 
fessor E. C. Case "honored" him by naming the oldest (Mid-Carboniferous) species 
of the group E. raymondi. 


Upon my head they plac'd a fruitless crown, 
And put a barren sceptre in my gripe, 
Thence to be wrench'd with an unlineal hand, 
No son of mine succeeding. 

Macbeth, Act III, scene i 

Dinosaurs have long been the spectacular assets of the paleontologist. Many a 
student has elected a course in "Paleo" because he or she heard that it was "all about 
dinosaurs," and much money has been raised for scientific purposes by using these 
gigantic creatures as bait. Although they loom so large in the public eye, their place 
in the great scheme of evolution is a comparatively unimportant one. It was their 
fate to make a triumphant, awe-inspiring march up a blind path, only to perish at 
the height of what seemed to be a career of ever-increasing prosperity. As we learn 
their history we see that the dinosaurs were just another branch of the reptiles, in- 
teresting because they were the chief terrestrial animals of the Mesozoic. 

The greater part of our knowledge of them has been obtained from the wonderful 
skeletons collected in the western parts of the United States and Canada during the 
past fifty years. Footprints and scattered bones in the Weald of southern England 
attracted attention as early as the middle of the eighteenth century and were de- 
scribed by G. A. Mantell in his Medals of Creation. A great English paleontologist, 
Sir Richard Owen, was the first to point out that these fragmentary traces repre- 
sented a group of enormous reptiles, named by him Deinosauria, the terrible lizards. 
The first specimens, however, were too incomplete to allow a full understanding of 
the creatures. Since Cuvier's law of "correlation of parts" could be applied with only 
partial success, the early attempts at making restorations resulted in what now seem 
to be mere caricatures. 

In America, as in England, attention was first attracted to dinosaurs by the dis- 
covery of tracks, which are abundant on the surfaces of certain layers of the Upper 
Triassic red sandstone in the Connecticut valley. Such impressions have been known 
for a long time, the earliest record being that of a specimen plowed up by Pliny 
Moody on his farm at South Hadley, Massachusetts, in 1802, but it was not until 
1836 that the tracks came to the attention ot the scientific world. In that year Dr. 
James Deane of Greenfield pointed out to Professor Edward Hitchcock of Amherst 
the birdlike footprints to be seen on the flagstones which were being used for side- 
walks in Greenfield. Hitchcock soon convinced himself that the impressions had 


been made by birds. He not only described them as such in the American Journal 
of Science but at once set about the task of obtaining all possible specimens, thus 
building up the unrivaled collection now to be seen in the museum of Amherst College 
(Fig. i, at right). Hitchcock founded a new science, ichnology, the interpretation 
of tracks, and was even able to persuade the state to publish two excellently illustrated 
boqks on the subject, the sole occasion, so far as I know, on which the Commonwealth 
of Massachusetts has taken cognizance of the existence of such an abstruse study as 
paleontology. Dr. Deane also published a book about the tracks, likewise in the belief 
that they had been made by birds. 

Although it is now known that there is little, if any, possibility that the foot- 
prints were left by winged animals, it is not surprising that they should have been 
so interpreted. The three-toed tracks are so obviously birdlike that a popular explana- 
tion of them as the footprints of Noah's raven was current before they were studied 
by scientists. The likeness is, in fact, more than superficial. Nat only do most of the 
tracks have three toes, but many show the imprint of the tip of a fourth, behind the 
others. Moreover, well-defined ones show the segmentation of the foot, which is like 
that of a bird. Another resemblance lies in the fact that the footprints of one of these 
trails follow each other in a single line, instead of being in two lines as is customary 
with four-footed reptiles, whose legs are far apart at the sides of the body. 

How, then, has it been determined that the tracks were made by dinosaurs rather 
than birds? It is chiefly an inference, based upon increased knowledge of the dino- 
saurs and the fact that so far no remains of birds have been found in any strata older 
than the Upper Jurassic. The study of the tracks themselves, however, has furnished 
some evidence against the original theory. Some trails are accompanied by a furrow 
that could have been made only by a dragging tail, which suggests a reptile rather 
than a bird. Others show an association of four- or five-toed impressions with the 
three-toed tracks, which is likewise indicative of a tetrapod. Moreover, many searches 
and a rather careful watch of the quarries in the Connecticut valley have resulted in 
the discovery of a few skeletons, all of reptilian creatures, most of them dinosaurs. 
Many dinosaurs walked on their hind legs in a birdlike fashion, and many of them 
had but three functional toes on the hind legs. Since the few skeletons which have 
been found are of this type, one naturally infers that the tracks were made by such 
creatures rather than by birds. 

If attention be turned from tracks to skeletal remains, we find that these animals 
constitute two great groups, now generally recognized as distinct orders of reptiles 
and distinguished from each other principally by their teeth and by the arrangement 
of the pelvic bones. There are, however, certain characteristics common to all which 
have been united under the general term "dinosaur." All are diapsid reptiles; that is, 
the skull has two pairs of openings in the temporal region, one pair dorsal, one lateral. 
This same condition is well shown in modern alligators and crocodiles, the living 


animals most nearly related. Dinosaurs differ strikingly from crocodiles, however, 
in that many of them were bipedal and that all of them walked with the legs beneath 
rather than alongside the body, in a mammal-like fashion. Furthermore, few had as 
much defensive armor as a crocodile, the skin of the majority being naked or covered 
with thin scales. 

The first group of dinosaurs includes those with sharp, simple teeth, which in- 
dicate that they were carnivorous or descended from carnivores. To these the name 
Saurischia reptile-hipped has been given, because theirs is a normal reptilian 
pelvis, each half made up of three bones (Fig. 71, at left). The ilia are attached to 
from three to eight united sacral vertebrae; a pair of convergent ischia extend down- 
ward and backward; and simple pubic bones, directed forward and inward, unite 
at their distal ends in a broad symphysis. This group contains two suborders, the 
true carnivores, or Theropoda, and the great four-footed amphibious dinosaurs, the 
Sauropoda. * 

The second order is known as the Ornithischia, since its members had somewhat 
more birdlike hips (Fig. 71, at right). The pelvis differs from that of the Saurischia 
in that it appears to be made up of four pairs of elements, instead of three. It is some- 
times spoken of as tetraradiate, in contrast to the more common triradiate reptilian 
type. There are really only three pairs of bones, but each pubis has two branches, a 
prepubis extending forward, and a postpubis projecting backward parallel to and be- 
neath the ischium. There is some, although a remote, similarity between this sort of 
pelvis and that of a bird. All the Ornithischia were herbivorous animals with rather 
long teeth adapted for cutting coarse vegetation. All of them exhibit an unusual char- 
acteristic in that the lower jaw has a beaklike bone in front of the tooth-bearing ramus. 
This is called a predentary; hence the older and more appropriate name of Predentata 
for this group. The anterior portion of the upper jaw is also toothless, the muzzle 
probably sheathed in horn, forming a turtlelike beak. There are three rather distinct 
suborders. The first are the Ornithopoda, the bird-footed kinds, herbivores which 
walked on their hind legs, as did the Theropoda. Next are the four-footed armored 
creatures, Stegosauria, best protected of all dinosaurs but rather short-lived, geologi- 
cally. Last are the Ceratopsia, the big-skulled, four-footed race which was latest to 
appear and in some respects most specialized of all. 

We shall follow the histories of the five groups, trying to find their relationships 
to one another and the particular adaptations of each. 

The true carnivorous dinosaurs, the theropods, represent the most vital stock of 
the saurischians. Most of the tracks in the Connecticut valley were made by mem- 
bers of this group, and almost all the skeletons so far found in Triassic rocks belong 
to it. As would be expected, their chief characteristics are their long, sharp, conical, 
or bladelike teeth, and the sharp claws on their toes, the impressions of which have 
been of great assistance in the interpretation of tracks. The skull is small, with large 


openings behind and in front of the eyes. The vertebrae are lightly constructed, the 
fore limbs definitely shorter than the hind. The number of toes varies considerably, 
the early forms having five on both front and hind feet, the latest only three or, more 
rarely, two fingers and four toes. No species is known with five functional toes on 
all the feet, at least one, and generally two, being short and more or less useless. Even 
tl\e oldest known members of the group appear to have walked habitually on the 
hind legs in a birdlike fashion, the long, stout, and somewhat rigid tail serving to 
balance the anterior part of the body. The numerous footprints found in the Con- 
necticut valley do not in any instance show evidence that the carnivores ever brought 
the fore feet to the ground. Although many skeletons of theropods have been de- 
scribed, enough to give at least a glimpse of the general lines of evolution in this 
group, it is not yet possible to outline any satisfactory phylogeny. 

FIG. 71. At left, pelvis of a saurischian dinosaur. 7, ilium; Pu, pubis; 
Is, ischium. At right, pelvis of an ornithischian dinosaur, with a posterior 
branch of the pubis below the ischium. From O. C. Marsh. 

Plateosaurus (Fig. 75) seems to be the most primitive carnivorous dinosaur so 
far discovered. It was a relatively large animal for its time (Upper Triassic), the 
skeletons varying from fifteen to nineteen feet in length. It has five digits on both 
hands and feet, but both the big and little toes of the foot are short, so that the median 
ones bear most of the weight. The arms are about two-thirds as long as the legs. The 
pubic bones are joined mesially and are greatly expanded and flattened at the union. 
Ventral ribs are present. Remains of this animal have been found in the Upper Tri- 
assic of Germany and France. An allied American genus is the somewhat smaller 
Anchisaurus, represented by some of the few remains which have been found in the 
Triassic of the Connecticut valley. 

Although Plateosaurus and Anchisaurus are themselves slender, agile animals, a 
contemporary of the latter in Triassic days in the Connecticut valley is of even lighter 
build. This is Podofesaurus holyol^ensis, described by Dr. Mignon Talbot from an 
almost complete skeleton, the most recent of the finds in our American "New Red 
Sandstone." It is a little creature, only about four feet long, nearly half the length 
supplied by the slender tail. The arms are short, the hands retaining but three fingers. 
All the bones are slender and of extremely light construction, indicating an active 


animal with a birdlike gait. This is one of the oldest of the slender theropods known 
as coelurosaurs. 

Other Triassic carnivores which should be mentioned are two small, slim forms 
somewhat allied to Podo thesaurus. One is Hallopus, one-time resident of Colorado, 
and the other Procompsognathus, whose remains were found in Wiirttemberg. Their 
chief claim to fame is that the calcaneum, or heel bone, is so much elongated as to 
suggest muscles in the back of the leg strong enough for leaping. That some dino- 
saurs had this kangaroo-like habit would be a natural deduction, but study of tracks 
and trails gives no evidence to support the idea that they ever departed from a sedate 
and even-balanced gait. 

Several theropods are known from the Jurassic, the most interesting, because 
best known, being Allosaurus, represented by a wonderful specimen from Wyoming 
now mounted in the American Museum of Natural History. Larger than any known 
Triassic dinosaur, Allosaurus is thirty-four feet long and a little over eight feet high 
at the hips. The tail is longer than the rest of the body; the flat articulations between 
the vertebrae indicate that it was not very flexible. It could not have been curled up 
or thrown over the back but probably stuck out rather stiffly, serving as a counterpoise 
to the anterior part of the body. The hind limbs are long, the bones straight, and the 
knees bent forward as in the mammals, not outward as in most reptiles. The arms 
are short, not long enough to be used in walking, but each hand has three long, sharp, 
recurved claws, effective for grasping and rending flesh. The skull is large and wide, 
with a long lower jaw, articulated at the extreme posterior end, so as to give the mouth 
a large gape. The teeth are of medium size, conical, but with sharp, somewhat 
compressed, cutting crowns. 

That Allosaurus preyed upon the big amphibious reptiles is shown by fragmen- 
tary skeletons of the latter found in the same strata. Vertebrae, particularly those 
of the tail, appear to have had their spines bitten off; others are scratched and scored. 
When a jaw of Allosaurus was compared with these marks, it was found that they 
agreed exactly, the spacing of the teeth corresponding with that of the scratches. 
Moreover, while the vertebrae of a Brontosaurus were being removed from the rock, 
numerous broken teeth of Allosaurus were found but no other traces of their skeletons, 
an indication that the teeth had been fractured while the animal was feeding. The 
specimens mounted in the American Museum combine the fragmentary and scored 
Brontosaurus skeleton with that of Allosaurus, posed above it in a feeding position. 
Whether Allosaurus successfully attacked such a huge animal, or whether it was a 
carrion feeder, devouring only individuals that had died through accident or from 
old age, is still a debated question. The sauropods were, however, so unprotected and 
dull of wit that it seems as though they must have fallen easy prey to large and active 

Allosaurus was the largest and most spectacular of Jurassic theropods, but, curi- 


ously, strata of about the same age in Germany have furnished the remains of the 
smallest known dinosaur, the diminutive Compsognathus^ no larger than a cat. So 
small is this skeleton, and so curved its attitude on its slab of Solenhofen limestone, 
that the suggestion has been made that it represents an embryo. Viviparous habits are 
not unknown among modern reptiles, and the inference that the dinosaurs may have 
brought forth their young in a well-developed condition is not illogical. The dis- 
covery that representatives of one of the most specialized dinosaur groups, the cera- 
topsians, were oviparous, seems, however, to indicate that this theory is untenable. 
There is, in fact, no evidence that the single known specimen of Compsognathus was 
not that of an adult animal. It is a slender, light-boned creature, possibly a resident of 
the Jurassic upland forests, where food may have been scarce enough to limit its growth. 
On the other hand, like many modern carnivores of small size, it may have been 
one of the creatures which exist upon the scraps which fall from the rich man's table. 
The theropods continued to prosper throughout the greater part of the Cretaceous, 
reaching their culmination in Tyrannosaurus only a short time before the dawn of 
the age of mammals. Parts of three skeletons of this greatest of carnivores have been 
found, two in central Montana and one in South Dakota. The largest, as mounted 
in the American Museum in New York, is nearly forty-eight feet long, the top of 
the head nineteen feet above the floor. The skull, although of light construction, is 
imposing, being about four feet in length: Armed with large, sharp, recurved teeth, 
the longest of which project five inches from the jaw, this is perhaps the most savage 
mouth the surface of the earth has ever known. The upper part of the skull has 
numerous openings bounded by narrow bones, and a brain case so disproportionately 
small that the low mentality of the great brute is readily apparent. The neck vertebrae 
are large and closely articulated, indicating great strength and little flexibility. The 
shoulder girdle and ribs are of light construction, but the pelvis is large. The heavy 
pubic bones are greatly expanded where they unite at the distal ends and furnished 
excellent support for the viscera, whereas the long slender ilia attached to several 
vertebrae welded together in the sacrum formed the basis of a pelvis well adapted 
to support the great animal in its semierect position. The bones of the arms are short 
and slender; only three fingers were functional, but they are armed with great re- 
curved claws which show that the fore limbs, although reduced, were by no means 
useless. The hind limbs are exceedingly strong, the articulation of the bones unusually 
close-knit for a reptile. The head of the femur has a large knob (trochanter), which 
fits into the acetabulum of the pelvis nearly at right angles to the shaft. The shaft 
bears various knobs and rugosities which indicate the places of attachment of large, 
strong muscles. The feet have three large digits, each tipped with a powerful recurved 
claw, and there is a fourth toe, which, although it did not reach the ground, is armed 
with a sharp spur evidently not without its function. Not only the teeth but the hind 
limbs of Tyrannosaurus could have been used as weapons of offense. To see two such 


animals locked in a firm embrace, using the hind legs after the manner of cats, roweling 
each other, would have been a stupefying spectacle. There can be no doubt that the 
tyrannosaurs were complete masters of their world; yet they existed for only a short 
time and probably were not numerous. 

A Canadian contemporary of this the greatest of all terrestrial carnivores was 
Gorgosaurus (Fig. 76), a little creature only twenty-nine feet long, but worth separate 
mention because it was even more specialized than Tyrannosaurus. Its diminutive 
arms have only two functional fingers, the smallest number yet found in any dinosaur. 

Several other large and truly terrible lizards have been found in Cretaceous strata, 
but the only other theropods to be mentioned here are small, slender ones somewhat 
like the Triassic Podo^esaurus and classified in the same group, the coelurosaurs. The 
best known of these is Struthiomimus y the "ostrich mimic," from Albertan Upper 
Cretaceous. It is of medium size, thirteen feet long, with slender bones, long neck 
and tail; it differs from most other known dinosaurs in lacking teeth of any sort, a 
birdlike horny beak probably having taken their place, although it is not preserved. 
The four-fingered hand is elongate, adapted for grasping and pulling down branches 
but useless in seizing active prey. The skull, posture, and hind limbs are remarkably 
like those of running birds, such as the ostriches, but the long tail proclaims Struthio- 
mimus a dinosaur. A very similar creature, armed, however, with slender conical 
teeth, is the Upper Jurassic Ornitholestes of Colorado. 

Glancing back for the moment over the theropods, we find that they existed 
from the Upper Triassic almost to the end of the Upper Cretaceous. The first of 
them still retained the typical five toes of hands and feet, although not all digits were 
fully developed. Throughout their history there was a constant tendency toward de- 
crease in size and length of the arms,- accompanied by reduction in the number of 
functional fingers to three, at least, and in one instance to two. The hind limbs be- 
came progressively longer and stronger. The outer (fifth) toe disappeared in some; 
the first was rotated so that it was in opposition to the rest v as in the birds, and became 
so short that only -rarely was it long enough to reach the ground. The theropods 
were all digitigrade; that is, they walked on the toes. The upper bones of the foot 
(metatarsals) came to lie closer and closer together, so that they acted more or less 
as a single unit, although they were not joined together. There seem to have been 
three or more distinct phyletic lines, but the collected skeletons are not yet sufficiently 
abundant to establish them. Plateosaurus, which retains five digits, appears to be 
near the stem from which all were derived, and also to be ancestral to or closely 
related to the ancestor of the next group, the Sauropoda. 

The sauropods, the most gigantic of all dinosaurs, were in some respects rather 
primitive. Not only did they walk humbly on all fours but they retained the five 
toes of the early reptiles. All had small heads, long necks and tails, large bodies, and 
stocky legs. 


Brontosaurus was a heavy-boned animal, sixty to seventy feet long, fifteen feet 
high at the hips, of such massive proportions that its weight has been estimated at 
as much as ninety tons, although a more reasonable figure is about one-third as great. 
The bones of the limbs are massive and solid, and indicate great strength. The feet are 
short; the distal bone in each of the outer toes is stubby, suggesting the presence of 
a .hoof. One or more of the inner toes bore claws. In contrast to the massive limbs 
and feet, the backbone, although strong, is of light construction. The centra of the 
cervical and anterior dorsal vertebrae are opisthocoelous; that is, each is convex at 
the front, concave behind. The remainder of the centra, except the distal caudals, 
have flat ends. The cervicals bear short neural spines but seem complicated because 
the hatchet-shaped ribs are more or less fused to them. The trunk vertebrae are most 
extraordinary, the complicated neural spines being tall and provided with various 
pairs of processes by which adjacent ones articulate with one another. Although some 

of the spines are as much as three feet high, they are nevertheless light, for all their 
elements are in the form of thin bony plates. Even the centra of the vertebrae are 
not solid but are deeply excavated at the sides. As someone has said, the sauropods 
were the first animals to discover the principle of angle-iron construction. All the 
cervical and trunk vertebrae are built in this way, but the caudals are more solid, 
without lateral cavities. As with many other vertebrates, the tail retains a primitive 
type of vertebra. The skull is small, of light construction, with small openings back 
of the eyes and larger ones in front of them. The teeth are long and slender, confined 
to the anterior parts of the jaws. Their tips are somewhat spatulate, with poorly de- 
veloped anterior and posterior cutting edges. For an animal of such size they seem 
utterly inadequate. 

Diplodocus (Fig. 72) was in many respects similar to Brontosaurus, although a 
longer and more slender animal. The best skeleton, that of Diplodocus carnegiei in 
Pittsburgh, is eighty-seven feet long, the greatest length known from any complete 
specimen of a dinosaur. All the bones are more slender than the corresponding ones 
of Brontosaurus, and the height at the hips is less by a foot or more. Both neck and 
tail are longer and slimmer, one of the most remarkable features of Diplodocus being 


the long whiplash at the distal end of the tail. The appearance of one of these animals 
partially submerged in water must have been uncannily snakelike. 

The hind limbs of both Brontosaurus and Diplodocus were stouter than the fore, 
and the same condition prevailed in several other genera of sauropods. There is, 
however, one group in this suborder in which the fore limbs were actually longer and 
stronger than the hind. These are the brachiosaurs, best known from specimens 
collected in eastern Africa but also represented by various bones found in the western 
United States. Though no complete specimen has been recovered, the present opinion 
seems to be that some of these animals reached a length of ninety feet. If, as paleon- 
tologists seem to believe, the brachiosaur could raise its head to the limit of its tre- 
mendous neck, it must have been the giraffe of dinosaurs, rearing to heights not even 
remotely attainable by any other creature which kept its feet on the ground. 

The oldest sauropods now known are those found in the Mid-Jurassic of 
England. They reached their maximum abundance in the Upper Jurassic, the age 
graced by the presence of Diplodocus, Brontosaurus , the brachiosaurs, and other huge 
animals. At that time they were especially well intrenched in the region which is 
now Colorado, Wyoming, and Utah, and in East Africa. Few remains of sauro- 
pods, a single species from Maryland being the most important, have been found 
in the North American Cretaceous, but recently several large forms have been de- 
scribed from Upper Cretaceous strata in South America, and others have been reported 
from Africa, India, and Australia. Were it not that C. W. Gilmore has found a 
shoulder blade nearly five and a half feet long in the Upper Cretaceous of New Mexico, 
we should be inclined to believe that the sauropods were banished to the southern 
hemisphere after Jurassic times. 

Although skeletons of ornithischian dinosaurs are much more common in mu- 
seums than those of sauropods and theropods, their evolution is less fully known, for 
Triassic representatives of the order are so few as to afford but meager information 
about them. That the connecting links between the three known groups are missing 
does not surprise one, for nearly all the material has been derived from strata of 
Upper Jurassic and Cretaceous age. 

The least specialized of the Ornithischia are the Ornithopoda, or bird-footed dino- 
saurs. Misnomers seem to reach their culmination in the name of this group, for 
the Ornithischia do not have a really birdlike pelvis and the foot of the ornithopod 
is less birdlike than that of a theropod. The ornithopods were actually birdlike only 
in their habit of walking on their hind legs, a characteristic shared by the theropods. 
The oldest evidence of the existence of this group consists of tracks on the Upper 
Triassic sandstones in the Connecticut valley. They are identified as belonging to 
the ornithopods rather than the theropods because the digits terminate in blunt, hoof- 
like toes, not in claws. These oldest members of the group had four functional toes 
and a vestigial one (the outer one) on the hind feet and five fingers on the hand, 


the two outer ones reduced in size. Unlike the Triassic theropods, they were not at 
that time entirely bipedal but upon occasion brought the fore feet to the ground. 
Skeletal material from the Triassic is highly unsatisfactory. The small and incomplete 
remains of Nanosaurus, long known from the supposed Triassic at Canyon City, 
Colorado, show few primitive characteristics. In all likelihood the few remains found 
in the Upper Triassic of South Africa will eventually prove to be more useful in 
tracing the ancestry of the group. 

Neither are Jurassic ornithopods any too well known. Camptosaurus (Fig. 77), 
an animal from ten to seventeen feet long, possessing a small head, short neck, rather 
long tail, short arms, and long, slender legs, is represented by fairly complete material 
from the Morrison beds of Colorado and neighboring states. The hands retain five 
fingers, and the sacral bones are not coossified, both primitive characteristics. The feet 
have four toes, the fifth absent, the first reduced, so that the functional ones are the 
second, third, and fourth. The general appearance of Camptosaurus was much the 
same as that of the contemporaneous carnivores. According to Gilmore, the front 
limbs, although much reduced, were probably still useful in locomotion; he has 
therefore mounted the specimen in the National Museum at Washington in a 
quadrupedal position. 

The Cretaceous was the heyday of the ornithopods; at least, so it seems to us 
because of the numerous remains which have been found, but every skeleton repre- 
sents an individual tragedy. Perhaps for the ornithopods it was a time of "war, pesti- 
lence, and famine," when the strong and ingenuous perished and the weak and 
cunning survived. 

A striking ornithopod is Iguanodon, one of the few European dinosaurs repre- 
sented by complete material. Trapped in fissures in the Carboniferous rocks of Bel- 
gium while they roamed that country in Lower Cretaceous times, the skeletons of these 
creatures were buried by the sands of the encroaching sea. Miners, following coal 
seams, have penetrated the filled crevasses and recovered the remains of many. Sev- 
eral are now mounted in the museum at Brussels. From twenty to thirty feet long, 
Iguanodon was a powerful animal, with a heavy, stocky. skeleton. Locomotion appears 
to have been largely bipedal, although the front limbs were half as long as the hind, 
more powerful than those of most bipedal dinosaurs. Five fingers were present, the 
third and fourth slender, the first reduced to a single spikelike segment. This thumb, 
which extends outward at right angles to the axis of the hand, was a powerful spur- 
like instrument, probably used as a weapon of offense. A none too gentle nudge with 
it may have warned* friend or foe to keep a respectful distance. The skull was laterally 
compressed, with deep, powerful jaws armed with closely set, serrated teeth arranged 
in a single row on each side. The feet were similar to those of Camptosaurus^ but both 
first and fifth digits were absent. 

Ornithopods have also been found in southern England and the Isle of Wight. 


The latter locality has furnished the skeleton of the slender Hypsilophodon. It is of 
interest because it shows that there were light, slender herbivores as well as carni- 
vores. Furthermore, it is one of the few with teeth in the premaxillaries. The anterior 
teeth are simple, with narrow, sharp crowns. The fact that they do not reach the 
median line, but leave room for a narrow beak, indicates that the ancestor had a full 
set like those of the carnivores. There appears to have been in all the ornithischians 
a general simplification of the teeth in the modification for a vegetable diet, for 
although much more numerous than those of the theropods they are more loosely 
attached to the jaws, in grooves rather than in sockets. Finally, Hypsilophodon is of 
interest because it may have been capable of climbing about in trees. If this is proved, 
it will be the only known instance of arboreal habits among the dinosaurs. Othenio 
Abel has pointed out that the long slender hands and feet, and especially the long 
arms, may have been an adaptation to climbing. The animal was small, only a little 
more than three feet long. 

One more family of ornithopods, the Upper Cretaceous trachodonts or duck- 
bills, remains to be discussed in some detail. Numerous skeletons of Trachodon 
have been collected, the most nearly complete of them in Wyoming, Montana, and 
South Dakota. About thirty feet long, the animals reached a height of nearly seven- 
teen feet when standing on their hind legs. The arms were only about one-sixth 
the length of the legs but were capable of supporting the individual in a quadrupedal 
position. There are three functional fingers, the first being absent and the fifth small. 
The feet have three toes, each ending in a broad, hooflike expansion. The skull is 
depressed, constricted in the middle but broadened at the front into a bill-like expan- 
sion which suggested the popular name for the family. Although the teeth were 
restricted to the posterior part of the jaws, they were extraordinarily numerous, 
many of them being in use at the same time, while replacement teeth were held in 
reserve beneath the gums. In many respects the dental succession reminds one of 
that which obtains in sharks and skates. Barnum Brown is authority for the state- 
ment that each jaw is provided with from forty-five to sixty vertical and ten to fourteen 
horizontal rows of teeth, making a total of more than two thousand altogether. 

Trachodon is perhaps most generally known because of the "mummies" found 
during the past twenty-five years. They are the remains of individuals which dried 
up instead of decaying or being devoured after death. After being throughly dried, 
they were buried, probably as the result of being caught by the rising waters of a 
flood. In the process of desiccation the skin shrank about the bones of the limbs and 
chest, collapsing over the visceral area. When it decayed, as it did later, it left its im- 
pression oh the matrix in which the animal was entombed. The skin was thin, 
covered with small scales, apparently horny in nature but unlike the familiar over- 
lapping ones of the snakes or the large horny and bony plates of the crocodiles. It is 
obvious that they were of no defensive value. Their reduced condition suggests an 


aquatic adaptation of the trachodoris, for scales tend to be lost by reptiles which live 
in the water. Several characteristics of the skeleton indicate that Trachodon was a 
good swimmer. The laterally compressed tail must have afforded an effective organ 
of locomotion, and the broad processes of bone on the inner posterior face of the 
femur indicate powerful tail-muscles similar to those of the crocodile. The general 
opinion is that Trachodon was amphibious in its habits, a dweller on the seashore, 
equally at home on land and in the water. Many remains are found in marine deposits, 
a circumstance which is unusual for dinosaurs. 

Strata of Upper Cretaceous age in Alberta have produced wonderfully preserved 
skeletons of duckbills allied to Trachodon, most of them smaller but nearly all more 
peculiar. Corythosaurus, with a remarkable bony crest reminding one of a cock's 
comb, is one of the most striking. A specimen showing part of the skin, tendons, and 

FIG. 73. At left, a restoration of the ornithopod, Iguanodon, after W. E. Swinton and Vernon 
Edwards. At right, the spiny-frilled ceratopsian, Styracosaurus. Restoration from various sources. 

even impressions of some of the muscles is to be seen in the American Museum. It 
has many features that suggest aquatic habits. The Royal Ontario Museum in Toronto 
also possesses a splendid representative of this animal and of Saurolophus, a duckbill 
with a tremendous recurved crest. 

Taken as a whole, the evolution of the ornithopods parallels that of the theropods 
to a remarkable degree. Both appear first in the Upper Triassic and reach their 
culmination in the Upper Cretaceous. In both there was reduction in the fore limbs, 
loss of fingers and toes, and increase in the size of the largest individuals. In all these 
changes, however, the theropods went further than the ornithopods. 

Theropods, sauropods, and ornithopods were all, so far as is known, without 
defensive protection. The theropods carried the war into the enemy's country, their 
teeth and activity being their best defense. The sauropods must have relied largely 
upon their bulk and their aquatic habitat for protection. The ornithopods, so far as 
can be judged, flourished most after they had learned to depend upon swimming as a 
method of escape. Throughout Jurassic and Cretaceous times, however, there existed 



a group of sluggish, small-brained, but somewhat heavily armed ornithischians 
known as the stegosaurians. Representatives of this group are neither common nor 
particularly large. 

The oldest of the stegosaurs, Scelidosaurus^ was about thirteen feet long. Like 
all the armored dinosaurs it walked on all fours. Its small head was held rather low, 
because the fore limbs were short. Its armor, which was not greatly developed, con- 
sisted of small dermal ossicles and tubercles in longitudinal rows. The skeleton shows 
a curious mixture of primitive and specialized characteristics. The hand had four 
fingers, the foot three functional toes; yet the centra of the vertebrae are amphicoelous, 
as in the earliest reptiles. Another curious feature is the absence of any enlargement 
for the spinal cord in the sacral region, a most un-dinosaur-like characteristic. Were 

FIG. 74. Stegosaurus. A restoration, after an outline sketch by R. S. Lull. 

it not, indeed, for the presence of the predentary bone, one would doubt if this were 
really an ornithischian. The specimens were found in the oldest Jurassic rocks at 
Charmouth in Dorsetshire, England. 

Stegosaurus (Fig. 74), the most bizarre of all dinosaurs, is well known from 
complete skeletons about twenty feet long, found many years ago in the Upper 
Jurassic of Colorado and Wyoming. The tiny skull, short front legs, and extraor- 
dinary bony plates embedded in the skin of the back combine to give this quadruped 
an aspect unlike that of any other animal. The broad plates, alternated in position 
in two rows, one on either side of the middle of the back, had their thickened bases 
embedded in what must have been a tough hide. Carried erect along the body, the 
thin edges upward, they seem to have been of little worth as a protection. Perhaps, 
like the bristling fur of an angry cat, they served rather to impress the enemy than 
to ward off his attack. Near the end of the short tail the broad plates give place to 
heavy clublike spines, two feet or more in length. These were obviously not designed 

FIG. 75. The Upper Triassic theropod, Platcosaurus, one of the most primi- 
tive dinosaurs known. Photograph by George Nelson of the specimen he 
mounted in the Museum of Comparative Zoology, Harvard University. It 
is about sixteen and a half feet long. 

FIG. 76. Gorgosaurus, the most specialized of the heavy-boned theropods. 
Skeleton about twenty-nine feet long. Photograph by courtesy of the American 
Museum of Natural History, New York City. 

FIG. 77. Two specimens of Camptosaurus, a primitive Jurassic ornithopod. 
The larger is seventeen feet long, the smaller almost ten feet. Photograph 
through courtesy of Charles W. Gilmore and the United States National 
Museum, Washington, D. C. 

"Fie. 78. Protoceratops, the hornless ceratopsian, with its eggs. Photograph 
by courtesy of the American Museum of Natural History, New York City, 
where this group is shown. 


for mere passive resistance. The swishing tail of a cornered and enraged stegosaur 
must have been a powerful deterrent to the advances of even Allosaurus, the mighty 
carnivore domiciled in the same region. Stegosaurus had a small head and the most 
diminutive of brains but was not an animal to be attacked with impunity if he saw 
his opponent in time to turn his back. 

Srnall-headed armored creatures, more or less allied to Stegosaurus^ are found in 
rocks of Upper Cretaceous age, the Belly River formation, chiefly in Alberta, and 
also in the Lance, the latest Cretaceous strata in Montana. Several genera are known, 
one of which, Palaeoscincus, has been dubbed, aptly, the "animated citadel." Most 
completely armored of all known dinosaurs, it was broad-headed, flat-backed, short- 
legged, dull-witted, slow-moving, almost completely encased in bone. Bands of 
stout thick plates alternating with irregularly arranged bony bosses covered the sides, 
particularly above the limbs, protecting the more vulnerable ventral surface. Seem- 
ingly these huge creatures were safe even from the gigantic Tyrannosaurus, but like 
all dinosaurs, armored and unarmored, they were on the verge of extinction at the 
very zenith of their powers. 

Most dinosaurs had great bulk, massive hips and legs, long necks and tails, but 
practically all that have been mentioned so far had small heads. It is the distinction 
of the ceratopsians, the latest of the predentates in order of appearance, that their 
heads were large not merely proportionally large, but actually the largest heads 
possessed by any terrestrial animal. Many skulls and some nearly complete skeletons 
have been collected from Middle and Upper Cretaceous strata in western United 
States and Canada. Until recently it had been thought that the ceratopsians were 
strictly American animals, but the discovery of a member of the group in the Gobi 
Desert has dispelled this misapprehension. The swampy lowlands east of the rising 
Rockies were, however, their chief habitat. 

Most of the ceratopsians had horns over the eyes, the feature which suggested 
their name, and a horn on the mid-line, above the nose. Pictures, models, and mounted 
skeletons have made Triceratops familiar to almost everyone. Between twenty and 
twenty-five feet long, with an estimated weight of about ten tons, it was an exceed- 
ingly compact, bulky animal. The neck and tail were short, a large part of the former 
covered by the great frill of bone at the back of the skull, which in some individuals 
occupied fully a third of the total length. Both fore and hind limbs were straight 
and massive, the five short toes at the front and the four at the back terminating in 
broad, hooflike, lingual phalanges. The fore limbs were considerably shorter than 
the hind, so that the greatest height was at the hips, as in nearly all quadrupedal 
dinosaurs. The impressions of blood vessels on the surface of the solid horns show 
that they were really horn cores, like those of modern cattle; in life, they must have 
had a thick covering, probably of horny material, which may have increased the 
length six inches to a foot. 


TriceratopSy from the Lance formation of the uppermost Cretaceous, is one of 
the last of the dinosaurs. It is therefore instructive to look at Monoclonius, one of the 
ceratopsians from the Belly River formation (Mid-Upper Cretaceous) of Alberta. A 
splendid complete skeleton, seventeen feet long, is mounted in the American Museum 
of Natural History. The skull, about five feet in length, has a long horn above the 
nose but only a very short one over each eye, an arrangement that gives the animal 
much more the appearance of a rhinoceros than of a Triceratops. The frill of bone 
extending back over the. neck has a scalloped margin. It is not solid, but has two 
large lateral openings. The partially preserved skin of this specimen shows rather 
large polygonal plates surrounded by areas of smaller ones, an indication that the 
body of the ceratopsians was covered with a thick hide in which were embedded 
protective ossicles of considerable strength. 

Associated with Monoclonius are the remains of Styracosaurus (Fig. 73, at right), 
described by Laurence M. Lambe. Styracosaurus, although one of the most ancient 
ceratopsians, was one of the most spinescent. Not only did it have the long nasal 
and two short orbital horns of Monoclonius, but the margin of the frill bore long 
conical spines. The head was almost as spinose as that of the horned toad, a modern 
lizard of our southwestern states. 

Other genera are known, more or less closely allied to those which have been 
mentioned. Nearly all the earlier forms, from- the Mid-Upper Cretaceous, have a 
long nasal horn, whereas the horns above the orbits are rudimentary or short. On 
the other hand, all of the ceratopsians from the uppermost Cretaceous have long orbital 
horns and a short nasal one. In Diceratops, in fact, the latter has almost disappeared. 
This transformation suggests a change in the habits of the animals during the late 
Cretaceous times. 

The only ceratopsian yet known outside North America is the small hornless 
creature discovered by members of the third Mongolian expedition of the American 
Museum of Natural History, and named Protoceratops andrewsi by Walter Granger 
and W. K. Gregory. It is the female of this species which is supposed to have laid 
the famous dinosaur eggs, many of which were collected during the expeditions led 
by Dr. Roy Chapman Andrews (Fig. 78). More primitive than any American cera- 
topsian in its hornless skull, short frill, and small size, the adult being only about 
seven feet long, it was at first hailed as the ancestor of the whole group. Unfortunately, 
intermediate forms between the Mongolian and American representatives have not 
yet been found, and the late Cretaceous age of the strata from which the Asiatic 
specimens were obtained makes it seem doubtful that Protoceratops is actually 
ancestral. It is, however, very like the theoretical ancestor. 

A few more words should be added about the peculiarities of the ceratopsians. 
They difler from the other predentates in having a predentary bone in the upper as 
well as the lower jaw. They are also the only known reptiles with double-rooted 


openings were far back, close to the eyes; they also are large, for these animals had 
no gills but remained breathers of air, like their terrestrial ancestors. A curious 
feature is the position of the pineal foramen on the median line at the juncture of the 
frontals and parietals. The vertebrae are simple biconcave disks, similar to those of 
fish. The neural arches are not united to the centra and articulate but feebly with 
one ^another, a reductive adaptation to the aquatic environment. Ichthyosaurian ver- 
tebrae of the trunk region are easily recognized by the pair of tubercles or depressions 
low on the sides, for the attachment of the double-headed ribs; the anterior caudals 
have single-headed ribs. 

The most remarkable parts of the skeleton are the limbs (Fig. 81, at right). They 
contain all the segments of the legs of terrestrial reptiles, being in no respect com- 
parable in their osteology to fins of fish; yet they are curiously modified for their 
finlike function. The humerus, radius, and ulna of the arm, the femur, tibia, and 
fibula of the leg, are all much shortened and flattened. Two rows of carpal and tarsal 
bones are present, all similarly flattened, and all so modified into polygonal shapes 
that they fit together like blocks in a mosaic. The phalanges, likewise, are flattened, 
the fingers and toes held close together in such a way that there is little or no chance 
for movement between adjacent digits. Most remarkable of all, some species show 
extra rows of fingers and toes; there are, indeed, extra bones along the sides of all 
the elements from upper arm to fingers. There may be as many as a hundred phal- 
anges, for the typical formula, 2,3,4,5,3, or 2 >3>4>5>4> ls ignored by these specialized rep- 
tiles. As many as eight or nine fingers or toes may be present. This is the really 
noteworthy feature of the ichthyosaur, for no other reptile, no amphibian, no bird, nor 
any mammal, barring such freaks as six-toed cats, has more than five digits. Reduction 
is common in various groups, but increase occurs here only. Whether it came about 
by bifurcation, as in freaks, or whether it is from the formation of bones de novo^ 
is not known. A possible explanation is that each appendage originally had five digits 
but that a cartilaginous border was formed on either side as the limbs came to be used 
as paddles. Such cartilages may have provided centers of ossification which led to the 
formation of false fingers and toes. 

Although the ichthyosaurs probably had terrestrial ancestors, it has not been 
possible to discover them. The oldest members of the group are those found in the 
Middle and Upper Triassic of Germany, Lombardy, Nevada, and California. These 
are somewhat less specialized than the true ichthyosaurs, since the legs are longer 
and the bones less flattened. The pelvis is larger, the skull relatively shorter, the teeth 
less numerous and set in sockets, like those of terrestrial reptiles. The vertebrae are 
longer, less like those of fish, more firmly articulated by processes of the neural arches. 
The posterior portion of the tail is not bent downward so sharply, a fact that indicates 
the presence of only a small tail fin. In some species, at least, it was not a caudal 
appendage comparable to that of the true ichthyosaurs but a small dorsal fin behind 


the pelvis. All these features indicate that the Triassic forms, commonly called mixo- 
saurs, were less fully adapted for aquatic life than the Jurassic and Cretaceous ichthyo- 
saurs, and hence more like the terrestrial ancestor. 

It has been suggested that the Lower Permian Mesosaurus (Fig. 80) of South 
Africa and South America may be one of the initial members of the line, but proof 
is lacking. Mesosaurus is interesting both because of its geographical distribution and 
because it is the oldest known reptile to show definitely aquatic adaptations. It is a 
small, elongate, slender animal, with a long skull containing numerous long, slender 
teeth on the margins of the jaws and smaller ones on the bones of the palate. The tail 
is long and laterally compressed, evidently an organ of locomotion, suggestive of that 
of some of the mixosaurs. The fact that the lower limb bones are short connotes aquatic 
life, but the chief proof of aquatic habits is seen in the greatly elongated little toe, which 
in the South African species has an extra phalanx. The first or "big" toe is the smallest, 
and the fifth or "little" toe is the big one, an indication, according to Williston, 
that the feet were webbed. In contradistinction to those of the ichthyosaurs, the hind 
legs are larger and more powerful than the anterior ones, and the neck has several 
vertebrae. These characteristics suggest the plesiosaurs rather than the ichthyosaurs. 
On the other hand, Dr. J. C. Merriam has shown that the locomotion of the Triassic 
mixosaurs was accomplished largely by the aid of the limbs and not by the tail. He 
found that their hind limbs were nearly as large as the anterior ones; consequently, 
he believes that the mixosaurs are in a sense intermediate between the mesosaurs and 
the ichthyosaurs. As a matter of fact, however, it is not possible to say that this sug- 
gested line really indicates the ancestry of the ichthyosaurs. 

The plesiosaurs (Fig. 81) were fully as remarkable in their aquatic adaptations 
as the ichthyosaurs, a comparison of the two groups showing particularly well the 
different results attained when similar animals use unlike methods in solving the 
same problem. The ancestors of both ichthyosaurs and plesiosaurs were terrestrial 
reptiles which were attracted to aquatic life by the food in the sea. Both ultimately 
came to feed upon fish, belemnites, and other swiftly moving animals, as well as 
the less elusive clams, so that both had to become good swimmers. The ichthyosaurs 
used the tail as an organ of locomotion; the plesiosaurs probably swam in the manner 
of a sea turtle. 

The outward form of a plesiosaur is but little like that of an ichthyosaur. The 
head in most cases is relatively small, and there is a distinct neck, short in some 
species but abnormally long in others. The body is broad, somewhat flattened, the 
tail short, without a fin. As in the ichthyosaurs, the limbs are paddlelike, the fingers 
closely appressed and enclosed in a thick integument. But there are no extra rows 
of digits, and the hind limbs are longer and stronger than the anterior ones. The 
long-necked forms were perhaps the most remarkable, for no other animals show 
so great an increase in the number of cervical vertebrae. Giraffes and camels have 


long necks, but both have the typical seven cervicals of the mammals; the great 
length is due to the elongation of the individual bones, not to the introduction of new 
ones. The length of the necks of some birds is due both to an increase in the number 
and to the elongation of the vertebrae, although no bird has more than twenty-one 
cervicals. But some plesiosaurs have the astonishing number of seventy-six bones in 
this region. One had a head two feet long, a neck twenty-three feet long, the body 
nine, and the tail seven, the neck being more than half the total length. In spite of 
this great expanse, plesiosaurs (Fig. 82) apparently did not have the gracefully 
curved, swanlike forms usually shown in restorations. The cervical vertebrae articu- 
lated by nearly flat surfaces, allowing for little curvature. The head must have been 
carried rather stiffly. 

Although they were well provided with teeth, plesiosaurs appear to have had 
the birdlike habit of swallowing their food whole. This placed the burden of masti- 
cation upon the stomach, which, like the gizzard of the bird, seems to have been a 
thick-walled organ containing gravel and pebbles which aided in the trituration of 
food. Specimens have been found with as much as a peck of pebbles within the ribs, 
fragments foreign to the strata in which the fossils are found. Modern crocodiles 
have a similar habit of picking up stones to assist in grinding their food. 

Looking for a moment at the osteology, we find that the skull is in many respects 
similar to that of the ichthyosaur. The elongate, recurved, conical teeth are not so 
numerous, and are set in sockets, not in grooves. There is a single pair of dorsal 
temporal openings adjacent to the parietals, and the narial openings are close to the 
orbits. The vertebrae are similar to those of the ichthyosaur, but the centra are only 
slightly biconcave and thus less fishlike. The pectoral and pelvic girdles are both 
large and flattened, with platelike elements that gave firm support to the paddles. 
A rather striking feature is the presence of numerous ventral ribs, which made a 
sort of basket for the support of the organs between the shoulder and pelvic girdles. 
All the bones of the appendages are somewhat flattened, but the humerus and femur 
are long. The lower arm and leg bones are short, however, as are the elements of the 
wrist and ankle. All the fingers are elongate, the individual phalanges being long 
and rounded, not flattened as in the ichthyosaurs. The phalangeal formula is not 
constant, but each digit has supernumerary bones, the third and fourth being the 
longest, with, in some cases, as many as nine segments. 

The oldest plesiosaurs are represented by fragments from the Rhaetic, the youngest 
Triassic deposits. All the good specimens are from Jurassic and Cretaceous strata. 
The best of them are found in the lower Jurassic of southern and eastern England 
and Wiirttemberg, Germany, at all these localities associated with ichthyosaurs, and 
in the Cretaceous of various parts of the world. Small ones are from eight to ten 
feet long; the largest, from Kansas and Australia, are as much as fifty feet. 

Although the ancestry cannot yet be traced in detail, there seems to be no doubt 


that the plesiosaurs were closely related to the Triassic nothosaurs. Skeletons of the 
latter are found in Germany, Switzerland, and northern Italy, chiefly in marine 
strata of Mid-Triassic age. They appear to have been relatively small creatures living 
on the shores of the great inland sea which at that time covered a large portion of 
central Europe. They resemble plesiosaurs in the structure of the skull, the elongate 
neck, the stout pectoral and pelvic girdles, and slender limbs, with short lower arm 
and leg bones. The legs are not modified as paddles, nor are there extra phalanges. 
The nothosaurs appear to have been amphibious in their habits, showing a tendency 
toward aquatic adaptation chiefly in the shortening of the bones of the lower parts 
of the limbs. No species which could have been directly ancestral to any plesiosaur 
has yet been found. 

From time immemorial sea serpents have held a large place in the folklore of 
seafaring nations. Even now scientists dare not deny absolutely that large marine 
snakes still exist in the oceans. There is no tangible proof that they do; it is not even 
probable; yet a half century ago scientists would have said that the existence in modern 
seas of squids fifty feet long was not probable. Someone may yet catch a sea serpent. 
It is known, however, that sea serpents were common in the Upper Cretaceous oceans, 
for their remains have been found in some abundance in Holland, Belgium, and 
Kansas, and there are scattered records from various states of North America from 
New Jersey to Alabama and from North Dakota to Texas, from France, Germany, 
and far-off New Zealand. These, however, were sea serpents with legs, not true 
snakes, although closely allied to them. The best specimens are found in Kansas, 
but the group takes its name, mosasaurs, from the river Meuse (Maes, Maas). 

The mosasaurs (Fig. 83) are elongate, round-bodied, short-legged reptiles, fully 
adapted for marine life. About twenty-five species, representing several genera, are 
known. The shortest of them is eight feet long, the longest complete specimen thirty 
feet, although larger incomplete remains indicate a maximum length of more than 
forty feet. The body is covered with small overlapping scales, in contrast to the ichthy- 
osaurs and plesiosaurs, which probably were naked. 

The snakelike characteristics are seen particularly in the structure of the skull. 
One of the most striking features of the snake's skull shared, however, by that of 
lizards is the mobility of the quadrate bone, which is held to the skull by liga- 
ments only. This gives great freedom of movement at the back of the jaw and assists 
in the swallowing of large objects. The quadrate of the mosasaur, although larger 
than that of a snake, was equally free to move. As in the snakes, the two rami of the 
lower jaws were united at the front by extensible ligaments, not by coossification. This 
was another help in engorging huge mouthfuls, but not of so great assistance as was 
an elbowlike joint about the middle of each ramus, a feature unknown in other sorts 
of animals. The jaws were set with recurved teeth, two double rows above and two 
single rows below, exactly as in snakes. One can well imagine a mosasaur catching a 



huge fish, turning the elbows of the jaws inward to spread the front apart and allow 
a wide bite, then turning them outward to pull the prey in, till constant repetition of 
the movements forced the unwilling victim within the power of the swallowing 
muscles of the throat. 

The centra of the vertebrae are procoelous, that is, concave at the anterior end, 
but they lack the complex processes characteristic of snakes. The pectoral girdle is 
strong, but that of the pelvis weak, not attached to the backbone. Evidently the 
limbs were used as fins, not as organs of propulsion, the long flattened tail taking 
over that function. So far as is known, the tail had no elaborate fin, although in 
some species the neural spines are elongated in the caudal region. A wriggling, eel- 
like method of swimming seems to be indicated. The upper and lower arm and leg 

FIG. 83. Skeleton and skull of the Kansan Cretaceous mosasaur, Clidastes. 
The skeleton is 11.3 feet long, the skull 2.5 feet. From S. W. Williston. 

bones are short and flat. All five digits are present, but there is a variable number 
of phalanges. Instead of holding the fingers and toes close together, these animals 
spread them apart, producing a flexible fin like that of the fish rather than a stiff 
paddle. All the digits were doubtless enclosed in a continuous membrane, although 
this has not been seen. 

As with the groups described above, the ancestry of the mosasaurs is still in 
doubt. Some believe them to be descended from the aigialosaurs, whose remains 
have so far been found only in the Lower Cretaceous strata of Dalmatia. In their 
characteristics these animals are almost exactly intermediate between the mosasaurs 
and terrestrial lizards like the modern monitors. The skull resembles that of a mosa- 
saur, including so striking a feature as the hinge in the lower jaw. The pelvis, how- 
ever, is firmly attached to the backbone, and the legs are long and lizardlike, although 
the feet must have been webbed, as is indicated by the clawless toes, a sign of at least 
partially aquatic habits. 

The modern monitors are elongate lizards belonging to the genus Varanus. 
Inhabitants of Africa, Asia, and Australia, they have recently been rather widely 


kept in zoological gardens in Europe and the United States. Their heads are long 
and pointed, and their jaws show a suggestion of the mosasauroid hinge. Most species 
are wholly terrestrial, but others are excellent swimmers, progressing by means of the 
compressed tail. Although fossils to tell us their geological history before the Upper 
Cretaceous are wanting, it is probable that animals much like the monitors existed in 
Jurassic times. Some of the ancestral forms appear to have been terrestrial, others semi- 
aquatic. From the latter may have arisen the Lower Cretaceous aigialosaurs, and from 
them, the Upper Cretaceous mosasaurs. 

The ichthyosaurs, plesiosaurs, and mosasaurs were the most conspicuous marine 
reptiles of the Mesozoic, but associated with them were members of two other groups 
which should be mentioned briefly. These are the Thalattosuchia, commonly called 
marine crocodiles, and the marine turtles. 

The first of these groups contains only a few members, known solely from skele- 
tons found in the Middle and Upper Jurassic of Europe. They differ from crocodiles 
in lacking a bony palate and in having biconcave vertebrae, two primitive character- 
istics. The head is crocodile-like, with a long, slender snout and numerous teeth. 
A wide ring of sclerotic plates surrounds the eye. The body is slender, the tail long, 
ending in a caudal fin whose lower lobe is supported by the backbone, as in the ich- 
thyosaurs. Contrary to the condition which obtains in that animal, the hind legs are 
longer and stronger than the anterior pair. The limbs form narrow paddles, some- 
what like those of plesiosaurs. From ten to twenty feet long, these creatures in some 
respects suggest overgrown tadpoles, the hind legs seeming rather superfluous. Their 
short geological range and limited geographical distribution mark them as un- 
successful competitors of ichthyosaurs and plesiosaurs, some of whose characteristics 
mingle in them in mongrel fashion. The absence of any of the bony armor so well 
developed in the true crocodiles deprived them of a needed protection. 

The marine turtles have the distinction of being the only "old salts" to survive 
from the Mesozoic to the present day. The carapace is incompletely ossified, the 
limbs are paddlelike. One of the oldest, the most completely specialized for marine 
life, and certainly the largest, is Archelon (Fig. 84), whose magnificent skeleton is 
one of the ornaments of the museum at Yale. Since the carapace is reduced to a 
row of marginals, there are numerous vacuities between its long, slender ribs. The 
massive plastron, on the other hand, is composed of great plates margined by long, 
finger like prongs. The hands and feet are large, the phalanges elongate, clawless, 
apparently good organs for use in swimming. The skull is three feet long, the neck 
short, not retractile, and the carapace six feet in length. Including the tail, the whole 
skeleton is about twelve feet long; the width 'across the outspread flippers is a little 
greater. Dr. G. R. Wieland, who discovered this huge "leatherback" in the Upper 
Cretaceous rocks of South Dakota, estimates its weight when alive as three tons. Its 
parrotlike beak suggests that it may have fed on shellfish, in which the Cretaceous 


seas bounded. Whether Archelon was in the ancestral line of the modern leather- 
back turtles is a disputed question. At any rate, it and its Cretaceous relatives, the 
modern leatherback (Dcrmochclys) , and the true sea turtles (Chelonidae) all show 
much the same sort of adaptation to marine life that we observed in the mosasaurs, 
that is, a lengthening and spreading of the digits. 

Marine and terrestrial reptiles alike reached their maxima in the Mesozoic. They 
received equally severe checks at the end of the era. What controlled their destiny is 
still a mystery. It would be interesting to know in what ways Mesozoic lands and 
seas were particularly favorable to reptiles and what could have happened at the 
end of the Mesozoic to affect all environments equally. 

FIG. 84. At left, the skull of a large modern marine turtle showing the 
simple cotylosaurian arrangement of the bones. From S. W. Williston, The 
Osteology of the Reptiles. At right, Archelon, a huge Cretaceous turtle. From 
G. R. Wieland. 

One theory in explanation of the decline and fall of the reptiles at the close of the 
Mesozoic, suggested by the all too prevalent habit of judging all other phenomena in 
the light of man's experience, compares racial history with the life story of the indi- 
vidual. Thus reptiles as a group were in a youthful stage at the beginning of the 
Mesozoic, endowed with abundant energy and fecundity. Having no competitors 
on the land, they increased and multiplied and peopled the earth. Though they were 
carnivorous at first, the abundant food led some of them to vegetarian habits. Before 
the end of the Mesozoic all reptilian phyla were ages old, and it may be that, like old 
men, they had lost their youthful vitality and fertility, and were no longer resistant 
to disease. Energy had run down; old age had overtaken the race. In their doddering 
senility some had lost part or all of their teeth; a few had actually grown spines, 
considered by some students a sure sign of approaching extinction. 

A plausible theory, but poorly supported by facts. Among other animals, both 
carnivorous and herbivorous dinosaurs increased constantly in size and numbers 
till the time of their disappearance. Even the sauropods, ousted from their ancestral 


homes, survived in the southern regions where they found refuge. It is true that 
some of the spinose forms, notably Stegosaurus, fell by the wayside, but since this 
happened millions of years before the extermination of the group it can hardly be 
accepted as a portent. Many stegosaurians continued to flourish. The ceratopsians 
are the only other dinosaurs which developed spines, but they appeared early in the 
history of that line, and were evanescent features, the last and largest of the animals 
having the fewest horns. When we turn to the marine reptiles, we find that their 
history is parallel, except that there is no development of spinosity. Throughout their 
era there was constant increase in size, numbers, and diversity. Strangely enough, a 
large group of aquatic forms, the mosasaurs, appeared during the Upper Cretaceous 
at about the same time as the terrestrial ceratopsians. There is no suggestion whatever 
of racial old age or loss of vitality. Reptiles of land and sea were still advancing and 
differentiating when the great catastrophe overtook them. 

What sorts survived? Among the marine forms, only the doddering old tooth- 
less turtles; of those on land, the slinking lizards, sphenodons, and snakes, and the 
semiaquatic crocodiles, a race which has never left its ancestral environment. 

So far this survey has not included the flying reptiles. Their history is not exactly 
parallel to that of the other sorts, except in so far as they reached their culmination 
in size during the Upper Cretaceous. Their greatest apparent diversity was attained 
during the Upper Jurassic, but since animals with the power of flight are seldom 
preserved as fossils it is not at all probable that their true distribution is known. There 
is every reason to believe that their story is the same as that of the other groups. 

In our speculations upon possible causes of the extinction of dinosaurs, various 
factors were considered, such as epidemic diseases, the devouring of eggs by mammals, 
and alteration of the vegetation through changes in climate resulting from mountain- 
building. Although any one, or all, of these happenings may have been important 
in the case of the dinosaurs, it is obvious that no one of them could originate a 
world-wide catastrophe affecting reptiles of land, sea, and air. Racial senescence 
having been ruled out, there remains only one guess that is, worldwide reduction 
in temperature. As is well known, cold-blooded reptiles thrive in tropical regions. 
The further northward or southward one proceeds from the present equatorial 
belt, the smaller and the less numerous are the reptiles, till one reaches a region 
where there are none. There is a definite northern and southern limit to reptilian 
distribution at the present day, but that limit undoubtedly is considerably further 
from the poles now than it was during Mesozoic times. The evidence for this is 
not wholly dependent upon the known occurrences of Mesozoic reptiles; it is sup- 
ported also by what is known of the contemporaneous plants and corals. 

"It is generally recognized that the end of the Mesozoic was a time of an unusually 
high stand of continents and restriction of shallow seas. Such conditions, at the end 
of the Proterozoic, in the Paleozoic, and in the Pleistocene, were accompanied by 


worldwide refrigeration and glaciation. It must be admitted, however, that there is 
little evidence for such a happening at the end of the Mesozoic. There were, it is 
true, mountain glaciers in the San Juans of Colorado during the Eocene, but these 
ice streams seem to have been local and relatively unimportant. In short, the evidence 
is confusing, for Eocene floras in general indicate that the tropical and subtropical 
belts were then wider than at the present day. Perhaps we can only finish rather 
weakly with a prediction that it will eventually be shown that sometime late in the 
Cretaceous or early in the Tertiary there was a period of general refrigeration during 
which the flying and the large terrestrial reptiles, all the marine reptiles except the 
turtles, all the ammonites, and the large sessile pelecypods were extinguished. 


He seems to be a man sprung from himself. 


A third of a century ago there was published in London a fascinating book, 
Dragons of the Air, H. G. Seeley's summary of his lifelong studies of those most 
extraordinary reptiles, the Pterosauria. The popular conception of reptiles is that 
they are sluggish creatures, prone to spend their lives basking in the sun, moving only 
when disturbed. On the other hand, flying suggests the acme of activity, constant 
motion, supreme vivacity. These estimates are, in the main, justified, but anyone who 
has witnessed the rushes of an alligator, the scamperings of a lizard, or the angry 
lashings of an infuriated snake realizes that reptiles may be extremely active. Flying 
reptiles are not, then, entirely anomalous, although since none exists at the present 
moment they seem a bit out of place in the general scheme of nature. 

One small group of the modern lizards, the little "dragons" {Draco) of Java, 
make some pretensions to consideration as members of the aerial fraternity. They 
are arboreal, as most flying vertebrates are, but they make no effort to fly, confining 
their activities to gliding from higher to lower branches. In this they are assisted by 
lateral extensions of the skin of the body, supported by flexible prolongations of the 
ribs. Such makeshifts, however, do not lead to true power of flight. This is accom- 
plished only if the fore limbs are provided with a patagium, that is, an expansion 
of the skin between bones or a series of feathers which may be moved as occasion 
demands. The patagium of the pterosaurs consisted of a thin but firm outgrowth 
of the skin, supported in front by the greatly elongated fourth finger of the hand, 
stretching thence to the tail and the proximal (femoral) segments of the legs. This 
wing lacked both avian feathers and reptilian scales; hence it closely resembled the 
wings of bats, though whether so "leathery" as one would infer from the common 
textbook statements is not clear. The impression, as preserved in the few known 
specimens, appears to be that of an exceedingly thin and featureless epidermis. 

Paleontologists are still amazed by the extraordinary history of the pterosaurs. 
Their sudden appearance, so far as the fossil record is concerned, in the upper Triassic 
of Europe is comparable only to their sudden extinction after a brief visit to North 
America in Mid-Upper Cretaceous times. 

Most of the best-preserved Jurassic specimens are small creatures, about the size 
of common bats; a few reached a length of eighteen inches. During the Cretaceous, 


however, there seems to have been considerable increase in size. The largest complete 
specimens are the pteranodons of the Kansan Mid-Upper Cretaceous, but fragments 
of similarly large individuals have been found at many places. In Upper Cretaceous 
strata in the vicinity of Cambridge, England, is a greensand which for a short period 
was extensively quarried for fertilizer. Although only a foot in thickness, it con- 
tained enormous numbers of coprolites, with a phosphatic composition which made 
them valuable. Associated with them were numerous fragmentary bones of ptero- 
saurs, some thousands of which were saved and preserved in various English museums. 
All are relatively large, and Seeley found among them some which seem to have 
belonged to animals with a wingspread of at least twenty feet. 

All pterosaurs had long, lightly built skulls, with large orbital, preorbital, and 
narial openings, the latter well back. The temporal vacuities of most are small, as is 
necessitated by the shortness of the region behind the eyes. The cranial cavity is 
small, enclosing a brain with birdlike arrangement of lobes but of diminutive size. 
The articulation of the jaws of some is peculiar, for the quadrates extend forward, 
bringing the back of the jaw below the eyes. The long, slender, curved teeth are. of 
the grasping type, but are directed forward instead of backward. They were not 
set in continuous series but were widely separated, the anterior ones longer than those 
further back in the mouth. This condition suggests that reduction in number and 
size was already in progress in early Jurassic times; one is therefore not surprised to 
find that some of the later Cretaceous representatives of the group are toothless. 

Numerous skeletons appear to indicate that the head was carried in a birdlike 
position, approximately at right angles to the vertebral column. Although the evidence 
for this deduction is so ample that such a posture is accepted as the normal one by 
most paleontologists, one should accept the statement with some caution. It is true 
that the heads of flightless birds are held at an angle of 90 or less with the axis of 
the neck, and volant birds when perching or walking show the same attitude. But 
birds in flight, with outstretched neck, raise the head, bringing it more or less into 
the axial trend of the body. If one looks critically at the specimens of pterosaurs in 
which the skull appears to be at right angles to the cervical vertebrae, one notices 
that most of them have broken necks, as indicated by an abrupt change in the axial 
direction of the cervicals near the skull. The rounded knob of the single occipital 
condyle must have allowed great freedom of movement of the head, but no more than 
that enjoyed by most other reptiles. The position of the skull was probably more 
dinosaurian than avian; if angles must be mentioned, it is probable that the axis 
of the skull was more nearly 55 than 90 off the direction of the body. 

The neck is relatively short, with seven large vertebrae whose broad neural 
spines afforded places for the attachment of strong muscles. The trunk also is short, 
with ribs on nearly all vertebrae. The sacrum contains from two to five vertebrae 
in addition to the typical reptilian pair; hence it is more nearly comparable to the 


dinosaurian than to the avian condition. The ischium is attached to the ilium, which 
is narrow but elongate. These bones share the acetabulum for the head of the femur, 
as in the crocodiles. The pubes are free bones without sutural connection with the 
other elements of the pelvis. The significance of this unusual condition has not yet 
been determined, but it should not be interpreted as indicative of the descent of 
pterosaurs from crocodiles. 

The limbs are of light construction, the bones hollow, with dense walls, much 
like those of birds. The pectoral girdle also appears at first glatice to be avian, for 
there is a large sternum or breastbone, a portion of the skeleton which is ossified in 
but few reptiles. Some pterosaurs have, as an anterior extension of this bone, a 
process like the keel of a bird of flight. Situated as it is, with its flattened surface at 
right angles to the body of the sternum, this anterior process must have afforded 
support to the pectoral muscles; hence it was functionally, if not morphologically, 
a keel. Its position, chiefly in front of the breastbone, is doubtless to be correlated 
with the structure of the wing. 

% There are two possible interpretations of the bones of the hand. The presence 
of four fingers is obvious. Three of these are of normal reptilian type, each provided 
with a terminal claw. The outermost is greatly elongated, many times as long as 
any of the inner ones. This finger is now considered to be the fourth, but it was for 
many years identified as the fifth, or little finger. Why the difference of opinion on 
this subject? 

If one examines any well-preserved specimen, one finds on the inner side of each 
arm a backward-pointing bone which seems to articulate with the carpals. The obvious 
interpretation is that it is a metacarpal, and the vestige of a thumb. If so, the clawed 
fingers were the second, third, and fourth, and the elongate one the fifth. It was, 
however, long ago suggested that the reflexed bone was not a metacarpal but the 
remnant of an imperfectly ossified tendon. If so, this "pteroid" bone has no homologue 
in the normal reptilian hand, and the first clawed finger is the thumb. According to 
Williston, the latter interpretation is the proper one, the evidence lying in the number 
of phalanges in each digit. Ignoring the inner, questionable element, the first clawed 
digit has two, the second, three, and the third, four phalanges. The greatly elongated 
finger also has four. The formula may therefore be written 0,2,3,4,4, or 2 3>4>4>- 
The typical reptilian formula for the hand is 2,3,4,5,3; hence it is obvious that 
the interpretation advocated by Williston implies less modification than the other, 
and so is the more plausible. It should not, however, be stated dogmatically that 
the interpretation now widely accepted is the correct one; it is merely the more 
probable one. Incidentally, it is that of Cuvier, the first paleontologist to describe a 

There are two types of pterosaurs, the short-tailed or Pterodactyloidea, commonly 
called pterodactyls, and the long-tailed or Rhamphorhynchoidea. 


Rhamphorhynchus (Fig. 86) is the best known of the long-tailed tribe. Several 
good specimens have been found at Solenhofen, the most complete of them the one 
now in the Peabody Museum at Yale. This individual retains the impression of a 
large part of each wing and of a small rudder or stabilizer at the tip of the tail. The 
rhamphorhynchoids are considered somewhat less specialized than other pterosaurs, 

FIG. 85. Sketches showing possible habits ot pterodactyls^ according to 
Abel. A, sleeping; B, awakening; C, walking; D, ready to glide; E, catching 
a fish. All redrawn, with slight modifications, from O. Abel. 

since an elongate tail is primitive, and the little toe, although shorter than the others, 
is less modified than that of any of the pterodactyls. Rhamphorhynchus and its 
allies are found only in the Jurassic of Europe. 

The history of the short-tailed pterodactyls is also chiefly European, entirely so 
during the Jurassic. Seeley stated that the group shows no evolution, but it is a 
far cry from the toothed Pterodactylus of sparrowlike proportions of the Upper 
Jurassic to the huge toothless Pteranodon of the Upper Cretaceous. What he had in 



mind seems to have been that the steps in the evolutionary change have not yet been 

The small pterodactyls of Solenhofen need no particular description. They differ 
from Rhamphorhynchus chiefly in having a short tail, a short fifth toe, a rather elon- 
gate skull and lower jaw, and smaller teeth, which are directed forward less con- 
spicuously. The Cretaceous Pteranodon (Fig. 87), however, deserves at least a 
paragraph, partly because it is the only well-known American pterosaur and partly 
because it is one of the greatest freaks of all time. Neither the anatomical knowledge 
of a Cuvier or an Owen nor the opium-inspired imagination of a De Quincey would 
have sufficed to predict the structure of this extraordinary reptile. The skull is the 

FIG. 86. Rhamphorhynchus, showing the long tail and portions of the wing 
membranes. Two-ninths natural size. From S. W. Williston, The Osteology 
of the Reptiles. 

nost remarkable feature, with its elongate toothless jaws balanced by an almost 
equally long backward extension of the posterior bones. Not one of Pteranodon^ 
urassic ancestors showed any tendency toward the production of such a crest. The 
;ye seems totally out of place, far behind the posterior ends of the lower jaws and 
ibove the lateral temporal openings. The skull is nearly twice as long as the vertebral 
:olumn, the neck as long as the inadequate and feeble body. The hind legs are long 
>ut slender, obviously ill-adapted for walking or perching. The fourth finger and 
ts metacarpal are so greatly elongated that the wingspread must have been from 
wenty to twenty-five feet. Despite its great size, Pteranodon probably weighed no 
nore than twenty-five pounds, according to Dr. George Eaton, who arrived at this 
result after weighing the exceedingly thin-walled bones of a nearly complete skeleton. 
There has been a great deal of discussion of the habits and possible habitat of 
:he pterosaurs, Their general similarity to birds, their large wings, hollow, air-filled 
Dones, and keeled sternum, together with other minor characteristics which have not 


been mentioned, suggest that they had considerable powers of flight. On the other 
hand, the wings are too large and awkward for efficient flapping, and the elongate 
anterior support makes them at once much less mobile than those of birds or bats 
and much less under control. Recent studies of the soaring action of birds and 
experimentation with motorless gliders suggest that the action of the flying reptiles 
was more in the nature of volplaning than of true flying. With light bodies and a 
great expanse of wing, animals could accomplish long flights by taking advantage of 
various currents in the atmosphere. 

Seeley thought that on land the pterosaurs were chiefly quadrupedal, as is shown 
by his numerous restorations, in which the animals are shown on all fours, the wings 

FIG. 87. Skeleton of Pteranodon, the most specialized of the pterosaurs. 
From George F. Eaton. 

folded up at the sides. This interpretation was doubtless inspired by the walking 
attitude of the barbastelle bat which he figured. It is a question, however, just how 
accurate this figure is. Most members of this group of flying mammals have the knee 
twisted about so that it points backward rather than forward and hence are no great 
pedestrians. Seeley also considered that bipedal locomotion was possible in the ptero- 
saurs, but Williston argued against this on the ground of the weakness of the hind 
legs and the slenderness of the toes. It should be pointed out, however, that the 
vestigial nature or absence of the fibula indicates that either the pterosaurs or their 
ancestors must have made good use of the hind legs. This bone does not tend to 
disappear in sluggish quadrupeds or bipeds. 

Abel, whose studies of the habits of extinct animals have been based upon wide 
knowledge of comparative anatomy, came to the conclusion that the pterosaurs were 
arboreal. Their clawed front and hind feet enabled them to climb, not perhaps with 


the agility of a squirrel but sufficiently well to get about. There is no evidence of 
the rotation of a first toe which would enable them to perch like a bird, but the long 
slender toes would serve to hold them in an inverted, batlike position. Folding their 
wings about them and pulling in their heads, they may have passed slumberous hours 
swaying in the breeze. Tired of one position, they had only to spread their wings and 
glide to another. A favorable breeze might tempt them to a long soaring flight. If 
food became scarce, a whole colony might glide away to settle elsewhere. Seeley de- 
duced from the abundance of their bones in the Cambridgeshire greensand that they 
were gregarious. 

What was their food? Seeley and many others have said fish, and there are 
numerous pictorial restorations showing a pterosaur swooping down with a fish, 
held crosswise, in its mouth. I cannot at the moment think of one in which the animal 
is swooping up. I have often wondered if the kindly creature was not on the point 
of putting a stranded fish back in the sea. 

Why the fish-motive ? Probably because remains of the flying reptiles have been 
found chiefly in marine deposits. But Seeley himself pointed out that practically all 
the pterosaur-bearing beds in England contain numerous remains of terrestrial ani- 
mals and plants. One of the deposits, the Wealden, is chiefly or entirely of fresh- 
water origin. Seeley interprets this as meaning that pterosaurs fished in rivers or in 
near-by seas. 

But why fish at all? More natural is the restoration showing Rhamphorhynchus 
pursuing a dragonfly, for some fairly large ones rest beside it at Solenhofen. It is 
doubtful if any animal ever climbed a tree primarily to catch fish. There are two 
good reasons for tree-climbing, one to get food, the other to escape enemies, and 
possibly a third, intellectual curiosity. 

ZaccheuSy he 

Did climb the tree, 

His Lord to see. 

But it is doubtful if this sort of curiosity was rampant in Mesozoic times. It is probable 
that the ancestors of pterosaurs and birds climbed trees originally because of one or 
both of the incentives first mentioned. Perhaps they were chased up and, having 
found food there, returned again; perhaps, having tasted the fruit upon branches 
within their reach, they pursued it onto higher levels. Where flowers and fruit are 
present, insects abound an excellent mixed diet, much enjoyed for a time by our 
own ancestors. If persisted in, however, it leads to the swallowing of small insects, 
seeds, and fruits whole. There is no particular need for mastication. Through disuse, 
the teeth become feeble, and some or all of them disappear, as they did in the 

It is interesting to compare the history of the loss of teeth by the pterosaurs with 


the same process in birds, although not much is known about it in either line. Appar- 
ently it began at the back of the jaws in the flying reptiles, at the front in the birds. 
The ancestral condition of neither group is known, but it is supposed that each was 
derived from a carnivore with small, closely set conical teeth on the margins of the 
upper and lower jaws. The oldest known representatives of both have rather widely 
spaced teeth, an indication that reduction had been going on for some time. As to 
the flying reptiles, let us quote from Seeley. "A Pterodactyle's teeth vary a good deal 
in appearance. The few large teeth in the front of the jaw of Dimorphodon (Lowest 
Jurassic), associated with the many small vertical teeth placed further backward, 
suggest that the taking of food may have been a process requiring leisure, since the 
hinder teeth adapted to mincing the animal's meat are extremely small." If we are 
to trust Seeley 's figures, which are the best known, the hinder teeth were not at all 
adapted for mincing meat. They were few and far between, and although they 
might have helped to break up a fruit, or husk a seed pulled off by the longer an- 
terior teeth, they would have been capable of nothing more than putting dents in 
"meat." Seeley continues, "In Pterodactylus (Upper Jurassic) they are .short and 
broad and few, placed for the most part toward the front of the jaws. Their lancet- 
shaped form indicates a shear-like action adapted to dividing flesh." Such carnivores 
as bother to masticate their food chew with the posterior (cheek) teeth, not the anterior 
ones, which are used for grasping and holding. 

Not to prolong the argument, it may be said that there are two cogent reasons 
for believing that pterosaurs were not primarily piscatorialists. The first is the nature 
of the teeth. The real fish-eaters do not chew their food. Either they swallow it 
whole or cut it into pieces which will pass down their gullets. Those whose food 
was primitively fish, and some in whom the habit is secondary, have sharp recurved 
teeth which help to hold and ingest the prey. The forwardly inclined teeth of the 
pterosaurs would not have helped to retain active, struggling animals but would 
have served well enough in the pulling of fruits and seeds. The second is that, 
although it is conceivable that the flying reptiles could have swooped down, snatched 
a fish which happened to be at or near the surface, and then have risen again, the 
action would have been extremely risky. The slightest mischance would have involved 
a forced landing, and they were about as well adapted for getting off again as a 
land plane. Doubtless they would have floated well, but how could they get any 
lifting power from those long, ungainly wings? A bat can rise from the ground, 
although there is a general belief to the contrary, but it seems improbable that a 
pterosaur could have got off the water. 

The relative abundance of the remains of pterosaurs in the quarries at Solen- 
hcfen is natural enough, for they contain many terrestrial insects and a couple of 
birds. The deposit was formed near shore, as is shown by the presence of the skeleton 
of a small dinosaur. The Kansan "chalk" from which remains of Pteranodon have 


been recovered was deposited much farther from the coast. However, remains of 
Pteranodon are extremely rare there; they probably represent a few unfortunate 
individuals, blown offshore and forced down in storms. Their bodies may have 
drifted long before the flesh had sufficiently decayed from their light, air-filled bones 
to allow their corpses to sink. They were too awkward, and carried too much sail, 
to be successful when emergencies arose. It is not at all surprising that the group 
became extinct. 


One never rises so high as when one does not know where one is going. 

Cromwell to M. Bellievre 

The chief glory of the reptiles lies not in what they themselves accomplished 
but in the fact that they gave rise to the two most important groups of modern 
vertebrates, the warm-blooded feathered birds and the warm-blooded hairy mammals. 
Like many obscure mothers, they achieved their distinction through their offspring. 

As has been pointed out repeatedly, the similarities between birds and reptiles 
are so numerous and so obvious that the appellation "feathered reptile" can be used 
with propriety. Modern mammals, except the monotremes, are not so obviously like 
the reptiles, and hence are commonly considered to be more widely removed from 
the ancestral stock, more specialized. If one stops to reflect, however, it will be 
realized that the wing of the bird is more highly modified than the limb of any 
mammal, and that other peculiarities, such as pneumatic bones, the curious vertebrae, 
and the synsacrum, combine to produce a skeleton much more specialized than that 
of the "average" mammal. Furthermore, in Triassic times the mammals were so 
like reptiles that it is possible to trace them to a particular group. On the other hand, 
the numerous reptilian characteristics of the birds do not suffice to lead one to the 
parental stock. They are the most specialized vertebrates and on anatomical grounds 
may be placed at the summit of the animal kingdom. 

Remains of birds, in strata older than the Pleistocene, are among the rarest of 
fossils; well-preserved specimens are the greatest of paleonto logical treasures. Until 
1861 none was known from strata older than the Tertiary, and, since those of the 
Tertiary are similar to modern ones, not much had been learned of bird ancestry 
up to that time. Then came, in a space of some sixteen years, a series of remarkable 
finds in the Jurassic and Cretaceous which served to bridge, in part at least, the gap 
between feathered and scaled animals. 

The oldest known bird is Archaeopteryx, whose name, meaning ancient feather, 
was given to the first specimen found. To describe a modern genus if one knew 
nothing more of it than the imprint of a single feather on a bit of hardened mud 
would be considered a very foolish proceeding, but it must be remembered that 
when this one was found in quarries in the Upper Jurassic limestone at Solenhofen 
it was the first trace of a bird Mesozoic strata had produced. Curiously enough, only 
a month after it had been described as Archaeopteryx lithographica, a nearly entire 


individual was discovered at the same locality. It was secured for the British Museum 
(Natural History). Although at first supposed to be of the same species as that 
which lost the original feather, it was later christened Archaeopteryx macrura. It 
lacks the skull but is otherwise nearly complete. In 1877 another specimen, the 
most nearly complete yet known, turned up at Solenhofen and was purchased for 
the Natural History Museum at Berlin. Long supposed to be an Archaeopteryx, 
recent study has convinced paleontologists that it belongs to another genus; hence 
it is now known as Archaeornis siemensi. Since 1877 nothing avian has been found 
at this locality, so the catalogue of Jurassic birds is a brief one, two relatively complete 
specimens and a feather. O. C. Marsh, it is true, described a "bird" from the Upper 
Jurassic of Wyoming, but the fragments are so poorly preserved that it cannot be 
proved that they are avian. 

As Alexander Wetmore points out, there are many reasons why birds should 
be rare as fossils. Epicurean man is not alone in his appreciation of a "warm bird," 
although it must be said to the credit of other animals that he alone desires the 
"cold bottle." Many birds succumb to the attacks of carnivorous animals; even those 
which die a natural death are eaten, unless speedily buried. Since the limb bones 
are hollow, carnivores crush them easily; only the thicker portions at the joints 
escape comminution. Small birds are entirely consumed, even the bones being 
digested. The brain is a choice tidbit, easily obtainable by crushing the thin-walled 
skull, and, the breast and viscera being equally desirable, little is left after the predator 
has lunched. It is not, then, surprising that most of the remains to be found in paleon- 
tological collections are ends of limbs, toe bones, and other scattered scraps. 

The skeleton differs in many respects from that of either a reptile or a mammal. 
The skull, which has large orbits, shows little trace of sutures between the constituent 
bones. The quadrate is movably articulated with the skull, not incorporated in it, and 
thus is "free," like that of mosasaurs and snakes. One of the most characteristic 
features is the bar which extends along the lower boundary from the quadrate to 
the maxillary. It is a slender bone made up of the quadratojugal, the jugal, and a 
posterior projection from the maxillary. A bone of such form is present in no other 
animal. On the other hand, the single occipital condyle is an obviously reptilian 
feature. The neck is extraordinarily flexible, as must have been observed by anyone 
who has watched the snaky writhings of geese or swans. This mobility is made pos- 
sible by the saddle-shaped articulations between the vertebrae and by the numerous 
processes to which muscles are attached. If you are unfortunate enough to be served 
with the neck of a fowl at dinner, your struggles to remove meat from it will convince 
you of the complexity of the bones. 

The rigidity of the trunk vertebrae is in striking contrast to the mobility of the 
cervicals. The vertebral column, from shoulder to sacrum, is a solid support for 
wings and legs. The dorsal vertebrae are more or less coalesced, so that little movement 


is possible between them, and the posterior ones are united with the long, broad 
ilia to form what is known as the synsacrum. The true sacrum has only two vertebrae, 
as in the reptiles, but from nine to twenty-one more are joined with them, the whole 
united with the pelvis to provide an unusually strong support for a body held in 
a semierect position. The ilia are long and deep, the ischia in most cases fused 
with them posteriorly, whereas the pubes are small bones lying below the ischia. They 
are vestigial, apparently turned far back of the normal position for such bones. The 
tail, which is short, is composed of few vertebrae, most of them fused together. 
The flat, slender ribs are characterized by prongs directed upward and backward. 
These "uncinate processes" are characteristic, but, strangely enough, similar out- 
growths are present on the ribs of some stegocephalian amphibians. The pectoral 
girdle is so constituted as to afford unusual support for the anterior appendages. Al- 
though it is not actually attached to the vertebral column, the muscular union is close. 
The shoulder blades project backward, parallel to the dorsal vertebrae. The collar- 
bones are united in front, forming the furcula ("wishbone"), and the coracoids join the 
anterior end of the breastbone. The latter is large, although thin. It is smooth in 
birds which have lost the power of flight but, as every carver knows, has a deep 
median keel in those with functional wings. The most powerful of the wing muscles 
are those connected with the sternum, and the presence of a keel adds greatly to the 
area to which they may be attached. 

The fore limbs have fewer bones than do those of most tetrapods. No vestiges of 
the fourth and fifth fingers remain, the wrist is incomplete, and the number of phal- 
anges much reduced from the typical reptilian formula. If we knew nothing of 
modern birds other than the skeleton, we should probably apply the term "de- 
generate" to limbs with so few bones; knowing as we do how powerful these organs 
are when equipped with feathers, we realize that they are really highly specialized. 
The hind legs are less modified, but still they have characteristics in which they differ 
from those of all other animals. The femur is short; the tibia and fibula are long. 
The latter bone is vestigial, in some cases joined to the tibia. The ankle joint is be- 
tween the two rows of tarsals, and no one of these is present as a free bone. Since those 
of the upper row are fused to the tibia, this combination is termed a tibiotarsus rather 
than a tibia. The lower tarsals are ankylosed to the upper ends of the metatarsals, 
which consequently are called the tarsometatarsals. The second, third, and fourth 
metatarsals are coossified into a single rather long bone. The first is vestigial, only 
the lower end remaining, and the fifth is absent. Most birds have four toes, the 
fifth (the little one) having been lost. The perching birds have the first toe behind, 
in opposition to the others. Many flightless birds have only three toes, the African 
ostrich but two. 

Since the Jurassic birds are the oldest known, they are of the utmost importance, 
and their remains have been repeatedly studied and described. Nevertheless, much 


is still to be learned about them, and more and better specimens are urgently needed 

if the problems which they raise are to be solved. 

An Archaeopteryx (Fig. 91) or an Archaeornis flattened in a slab of limestone 
covers a considerable area, but in life neither could have been much larger than a 
domestic pigeon. As preserved, the fossils retain impressions of the larger feathers of 
the wings, hind limbs, and tail, and the substance of most of the bones of the skeleton, 
although there are serious lacunae. Only the single specimen of Archaeornis retains 
the skull, which is flattened, badly crushed, and imperfect in the highly important 
posterolateral and temporal regions. In spite of these imperfections, it displays a 
large birdlike orbit, with remains of a ring of sclerotic plates, and a rank of thirteen 
small conical teeth, implanted in sockets, in the upper and presumably in the lower 
jaws. There is no beak. 

The neck appears to be relatively short, but since cervical and dorsal vertebrae 
are alike, no exact cervical formula can be stated. Most students have described the 
vertebrae as amphicoelous, which may be true, but the statement is an assumption, 
for the specimens have been considered too precious to permit sufficient dissection 
to show the true nature. Externally the vertebrae are simple, and it is certain that 
the articulations are not saddle-shaped, so the neck could hardly have been as flexible 
as that of a modern bird. Neither could the dorsals have been rigid. The sacrum is 
badly preserved, but the pelvic bones of Archaeopteryx have lately been revealed by 
delicate manipulation. The ilium proves to be so short that not more than four ver- 
tebrae could have been attached to it, a marked contrast to the number involved in 
the synsacrum of modern birds. The tail is long, with a pair of large feathers for 
each vertebra. Totally unlike the modern fan-shaped tail, it was constructed on the 
same plan as that of a lizard. The bones of the pectoral region are imperfectly pre- 
served. The scapula appears to have been slender, and a bone in front of it may 
represent the furcula. There is no trace of a sternum, although probably one was 
present. It may have been incompletely ossified or entirely cartilaginous. 

The pelvis is peculiar (Fig. 93 B). The ilium is short; the pubes are deflected 
downward and turned backward on the ventral side, the opposite ones in contact, 
although not actually united. The ischia, bones with irregular outlines, project down- 
ward and backward. Their conformation shows that the egg must have been small, 
with contents not more than one quarter the bulk of that of a modern bird of the 
same size. The wings, although feathered, are not much like modern ones (Fig. 88). 
The humerus is longer than the radius and ulna, and though there are only three 
digits, each is complete, tipped by a long, strong claw. Apparently the fingers func- 
tioned both as climbing and as flying organs. 

Until recently it was supposed that the hind limbs were like those of modern 
birds, but the removal of matrix which had partially concealed the bones shows 
that the fibula is as long as the tibia, although only a part of it is preserved, and 


that at least one of the tarsals, the fibulare, which is the fourth in the first row, is a 
free bone, not fused with the end of the tibia. Moreover, the median metatarsals are 
not joined to one another, although they lie very close together. Each probably has 
joined to its proximal end the corresponding tarsal of the lower row. The first 
metatarsal is behind the others and supports a short toe, opposed to three others, 
as in the perching birds. The outer, fifth, metatarsal is vestigial and has no toe. 

From this detailed description of the skeletal features of the Upper Jurassic speci- 
mens it is evident that they have more reptilian than avian characteristics. They 
were, as Gerhard Heilmann has said, "warm-blooded reptiles disguised as birds." 
Almost their only strictly avian characteristic is the presence of feathers. The list of 
reptilian features, however, is formidable. It includes both positive and negative 

FIG. 88. Wing of a modern pigeon, above, compared with that of Archae- 
ornis, below. H, humerus; R, radius; U, ulna; c, carpals; me, metacarpals; 
di, d2 y dj, first, second, and third fingers. From Gerhard Hermann's Origin 
of Birds, by permission of the D. Appleton-Century Company. 

features, such as the lack of a bill, the presence of teeth, a fixed rather than a free 
quadrate, simple vertebrae, lack of differentiation between cervicals and dorsals, 
few sacral vertebrae, a long tail, the absence of uncinate processes from the ribs 
(this may be due to poor preservation), the presence of ventral ribs, a humerus longer 
than the forearm, fingers with claws, a weak, possibly unossified sternum, a short 
ilium, a long fibula, at least one discrete tarsal, and separate metatarsals. 

Lower Cretaceous deposits have as yet yielded only scanty and unsatisfactory 
traces of birds, but considerable dispersal must have been taking place during this 
time, for avian fossils are widespread in Upper Cretaceous strata. Since most of the 
remains are exceedingly fragmentary, little is known about these latest Mesozoic 
representatives of the group. Only two genera are really well known. They are 
Ichthyornis and Hespcrornis, both described by Marsh from remarkable specimens 
found in the yellow chalky limestone of western Kansas. 

Ichthyornis (Fig. 92, at left) appears to have been about as large as a domestic 
pigeon. It was a true bird of flight, as is shown by the large, deeply keeled breastbone 


and well-developed wing. The feathers have not been found, but the ulna shows 
tubercles for the attachment of secondary wing feathers, and the ankylosed meta- 
carpals afforded a firm basis for the primaries, so it may be inferred that the wings 
had the same general arrangement of quill feathers as modern ones. Despite numerous 
primitive characteristics, the skeleton is surprisingly modern. The chief primitive 
features are teeth in sockets along the maxillaries of the upper jaw and the whole 
length of the dentaries of the lower; the two rami of the jaw not fused at the sym- 
physis; the vertebrae slightly biconcave; the three components of the pelvis not fused 
with one another at the posterior ends, and the small brain. In contrast to these are 
the following: the presence of a bill on the upper jaw, as indicated by the absence 
of teeth from the premaxillae; somewhat complex lateral and dorsal processes on the 
vertebrae, the third cervical foreshadowing saddle-shaped articulations; distinctly 
modern wings, with radius and ulna longer than the humerus; coossified metacarpals 
and reduced, clawless phalanges; birdlike scapulae, coracoids, and sternum; an elongate 
pelvis, attached lo a synsacrum; and completely avian hind limbs, with a vestigial 
first toe in opposition to the others, its small size indicating that, although descended 
from a perching type, Ichthyornis itself was not of that habit. 

The characteristics appearing in the second list are so much more important 
than those in the first that it may fairly be said that birds were about as highly spe- 
cialized for flight in Upper Cretaceous times as they are at present. They must have 
passed through the more important stages in their evolution between Upper Jurassic 
and Upper Cretaceous times. This has been cited as an instance of rapid evolution, 
but as a matter of fact the time that elapsed between the Upper Jurassic and the 
Upper Cretaceous was probably as long as that from the Upper Cretaceous to the 

In the same strata with Ichthyornis are found the remains of Hesperornis 
(Fig. 89), a large wingless creature fully equipped for aquatic life. Specialized as the 
birds of flight are, this aquatic animal superimposes specialization upon specialization. 
Crowded off the earth by overpopulation or by enemies, or lured therefrom by certain 
kinds of food, birds learned to fly and became adapted to life in the air. But no 
sooner had they mastered the new form of locomotion than some members of the 
group went back to live on land; others, still more adventurous, essayed life at sea. 
In the new environment, the wings, the perfection of which had been the crowning 
accomplishment of the race, became more or less worthless and through lack of use 
tended to atrophy. Everyone is familiar with the fact that the wings of many modern 
ground and aquatic birds are inadequate for sustained flight. Yet wings are present 
even in the flightless birds. Few if any of the modern Aves show so great a reduction 
of wing as Hesperornis. 

Comment has just been made on the suggestion that the perfection of the wing 
between Upper Jurassic and Upper Cretaceous times indicated rapid evolution, and 


a reason for discounting it has been shown. But Hesperornis tells a different story. 
Almost completely wingless, it was nevertheless descended from a bird of flight, 
indicating, if one may so express it, twice as much evolution as Ichthyornis. That is, 
full adaptation to life in the air was accomplished; then, secondarily, there was a 
complete change to a life of swimming and diving. Unfortunately, there are no data 
on rates of change in evolution. Despite certain opinions, based upon Biblical training, 
it is not at all certain that it was slow on the narrow upward path and rapid on the 
broad downward way. In fact, it is unlikely that there is any definite rate; progress 
one way or another depends upon many circumstances, partly inherent in the organ- 
ism, partly external. 

FIG. 89. Restoration of Hesperornis, redrawn after G. Heilmann. 

Specimens of Hesperornis, spread out on slabs of chalk, reach a length of five 
feet from tip to tip. Because of the peculiar construction of the legs it is difficult to 
say what its height was, but it was probably about as large as a swan, although the 
elongate head is un-swanlike. The arrangement of the teeth seems to be the same as 
in Ichthyornis, except that they were set in grooves instead of having separate 
sockets. The elongate cervical vertebrae are relatively simple but have saddle-shaped 
articulations. The ilia are long and connected with the many vertebrae of the 
synsacrum, an avian feature, but the long slender ischia and pubes are separate, 
forming a pelvis which as a whole is primitive. The ribs have uncinate processes; 
the sternum is long and well developed, although without a keel: the clavicles are 
not united to form a wishbone. The coracoid is short and wide, forming, with the 
long birdlike scapula, a firm support for a wing; but the wing is not there. The 
slender humerus dwindles to a point at the distal end, showing conclusively that it 
is only a vestige. It must have been without function. The hind limbs are well-de- 


veloped and strong, larger, in fact, than those of a perching bird. The femur is 
short, the tibiotarsus long, stout, and hollow, the tarsometatarsals long and strong 
but curiously articulated so as to turn outward at right angles to the body. As Lull 
has shown, this proves that the feet could not be brought beneath the body; hence 
Hesperornis could never have walked well on land but was doomed forever to a 
paddling existence. For such a life it was, however, well fitted, for it not only had 
a thick coat of downy feathers but the outer (fourth) toe is tremendously developed 
and fully twice as long as the third. The first does not turn backward as in the 
perching birds but forward, contributing its mite to the formation of a powerful, 
webbed foot; yet it is far from deserving the title of big toe, for it is vestigial, much 
smaller than the others. 

As may be seen by the sketchy description outlined above, the skeleton of Hes- 
perornis presents a queer combination of primitive (reptilian), advanced (bird of 
flight), retrograde (wingless), and aquatic characteristics. Those of the first two 
categories suggest the ancestry, those of the last two the habits of the creature. It 
was apparently a strong swimmer, a good diver, a fish-eater. Entirely efficient at sea, 
it could only have flopped about on land, if it could make any progress at all. Pos- 
sibly it laid its eggs in nests built of drifting vegetation, as the modern grebe does, 
although it is also possible that they may have been deposited in the sand, near high- 
tide mark. Wetmore suggests that Hesperornis could have floundered about on shore, 
progressing, as the seal does, on its breast. This does not seem likely, however, for the 
pectoral appendages were of much less use than those of seals. 

So much for Cretaceous birds. In some respects their characteristics are inter- 
mediate between those of Jurassic and modern ones, but so few genera and species 
are well known that it is impossible to trace family lines. It is not probable that any one 
of them is directly ancestral to any living animal. Yet these few representatives of the 
ancient avian fauna do help to reconstruct, partially at least, the chain of structural 
changes through which birds must have passed in the course of their evolution. 

Cenozoic birds are much more like modern ones than those of the Mesozoic 
are. Most of them can be referred to families represented in the modern fauna, 
although there are some peculiar creatures of doubtful affinities, probably repre- 
sentatives of extinct groups. Only the large flightless ones will find a place here. 

The most spectacular of all Eocene birds was one of the unexpected discoveries 
made by a party from the American Museum of Natural History in 1916. While col- 
lecting mammals in the Big Horn basin of northern Wyoming, they came upon the 
nearly complete skeleton of a huge bird which apparently belongs to the same genus 
as one previously known from fragmentary specimens from New Mexico, to which 
the name Diatryma (Fig. 90) had been given. About seven feet in height, it enjoys 
equal distinction as the largest North American bird and the oldest of all ground 
birds. It has no more remnants of wings than Hesperornis, no vestige being visible 


externally. The hind legs are long, strong, provided with three large toes. The neck 
is short, with large vertebrae, to which were attached the powerful muscles necessary 
for holding the most massive head ever seen on a bird. The skull is about seventeen 
inches long and six and one-half inches deep in front of the eyes. There are no teeth; 
nevertheless, one is inclined to believe that it was a carnivore, preying upon small 
animals of all sorts. Although without any particular organs of defense other than 
the strong hind legs and the powerful beak, Diatryma, when on the alert, could not 
have been in any great danger from the rather small contemporary carnivorous mam- 
mals. The general opinion seems to be that it was related to the cranes and rails, 
although not ancestral to them. 

FIG. 90. An Eocene flightless bird, Diatryma, and a contemporary horse, 
drawn to the same scale. The horse, about one foot high at the shoulder, 
is redrawn after Charles R. Knight, the bird after Matthew and Granger. 

The Miocene strata of Patagonia have produced some large and strange ground 
birds, larger even than Diatryma. One of the most remarkable is Phororhacos. Im- 
agine an ostrich with a head and neck as large as those of a horse, the skull bearing 
a great curved beak as sharp as an ice pick, and you have some idea of this peculiar 
creature. The largest known skull is about two feet long, exceedingly massive, but 
not so deep as that of Diatryma. The cervical vertebrae are enormous, five inches 
across, larger than those of any other known bird. Vestigial wings, totally inadequate 
as organs of flight, may have been of some assistance as offensive or defensive weapons. 
The legs are large, with three strong toes on each foot. One suspects that Diatryma 
was a carnivore: one is sure that Phororhacos was a bold aggressive carnivore, 
with little to fear and itself an object of terror to all inhabitants of the land. The 
huge legs suggest power in running, scratching, and fighting, so that, although it is 
considered to be a relative of the cranes, it is difficult to think of it as a wading bird. 
Phororhacos and Diatryma were related, doubtless through a common ancestor. 


The occidental world provided the first of the ground birds, but their greatest 
diversity was in the Orient during Pleistocene times. These eastern feathered creatures 
are not in any way related to those just described. 

The largest and most spectacular of all birds are the moas of the Pleistocene of 
New Zealand. Almost a hundred years ago missionaries to that southern Common- 
wealth heard from the natives stories of a monstrous human-headed creature which 
lived, guarded by two huge lizards, on a mountainside far in the interior. No man 
had seen this monster and lived to tell the tale, for on the approach of human beings 
it rushed forth and trampled them to death. Although none had actually seen it, 
the huge bones were well known. Instigated by these reports, travelers and naturalists 
set out in quest of what was locally called the moa. Although they never found it 
alive, they came upon its bones in the muck of swamps, in old lake beds, and in caves. 
Specimens were found retaining bits of skin and dried tendons, and old camp sites 
yielded charred bones and broken eggshells. Such evidence, coupled with tradition, 
indicates that the moas existed until recent times. It is probable, indeed, that the 
aborigines may have delivered the coup de grace to this tribe. Remains of many were 
sent to Europe in the middle of the last century, and from twenty to twenty-five 
species have been described, grouped under four or five generic names. Not all were 
giants, but all were relatively large. Many of them passed through the hands of the 
great anatomist, Sir Richard Owen, who fortunately had a relative in New Zealand 
to supply him with specimens which are now in the British Museum (Natural 

The larger moas (Dinornis and Palapteryx} are of extraordinarily massive con- 
struction. The head is relatively small, but broad, with a wide, short, pointed beak. 
The vertebrae are large, especially the dorsals; the breastbone large, convex or flat, 
but without any trace of a keel. A scapula is present, but there is no glenoid cavity 
at the junction with the coracoid, hence no articulation for a humerus, and no vestige 
of a wing. The pelvis is curiously primitive, the single archaic feature of these re- 
markable birds. The ischia are not united with the ilia, nor are the pubes in contact 
with the ischia. In this respect the hipbones are reminiscent of those of Ichthyornis 
and Hesperornis. Whether this is a truly primitive characteristic or a reversion to an 
ancestral status cannot be determined in the absence of representatives of the group 
older than the Pliocene. The question is of considerable importance, for if the con- 
dition is- due to secondary adaptation it constitutes an important exception to Dollo's 
"law of the irreversibility of evolution." The hind legs are "perfectly enormous." The 
femur is relatively long for a bird but not more than half as long as the tibiotarsus; 
both are of large diameter, greatly enlarged at the joints. Although vestigial at the 
lower end, the fibula is a big bone, mostly behind the tibia. Three large spreading 
toes are supported by the massive tarsometatarsal. 

The largest of the moas, Dinornis maximus, appears to have been at least ten 

- > v *vi,vfi' T v ,. .-foif^f '^ftiMfv ^ 'Jap--'. ^FI f^ 1/7 -f/." '''wpi'kji 

FIG. 91. The specimen of Archaeopteryx in the British Museum (Natural 
History). From B. Petronievics. 

FIG. 92. At left, a restoration of Ichthyornis. About one-third natural size. From O. C. Marsh. 
At right, a skeleton of Aepyornis. Original about ten feet high. From L. Monnier. 


feet in height opinions differ. Its proportions were those of an ostrich, but since it 
was much taller, was larger in the body, and had more massive bones in the legs, it 
must have been a far heavier creature. The power of the legs must have been tre- 
mendous. We may, with Dr. F. A. Lucas, put the question, "If a blow from an irate 
ostrich is sufficient to fell a man, what would be the kicking power of an able-bodied 

Madagascar, as well as New Zealand, was inhabited in Pleistocene and more 
recent times by huge birds. All those now known belong to the genus Aepyornis^ 
long famous for the size of its eggs. Eggs are notoriously fragile; yet in one way or 
another some escape destruction. Birds' eggs are known from rocks as old as the Cre- 
taceous, and collectors come upon them from time to time in various Tertiary strata. 
Aepyornis (Fig. 92, at right) eggs have been found chiefly in the muck of swamps, 
although the most nearly perfect ones are those which have been transported by streams 
from the place of their original burial. One of the best specimens is said to have been 
taken at sea. The largest known has as its greatest diameters nine by thirteen inches. 
The egg of an ostrich measures four and a half by six inches; hence the shell of the 
egg of Aepyornis would hold the contents of six ostrich eggs. If the comparison be 
carried further, it appears that the shell would hold the bulk of more than a gross 
of ordinary hens' eggs and of some thirty thousand hummingbirds' eggs. In plain 
terms, the capacity is about two gallons. 

Judged by the size of the egg, Aepyornis should have been about six times as 
large as an ostrich. But, as Lucas has pointed out, there is no definite relationship 
between the stature of a bird and the size of its egg. There still lives in New Zealand 
a flightless, four-toed bird, somewhat distantly related to the moas, famous for its 
eggs. The little Apteryx, smaller than our common barnyard fowl, produces an egg 
which measures three by five inches, and has about one-third the weight of its pro- 
ducer. Judged by the egg alone, Apteryx should be four feet high. But skeletons (one 
is on exhibition in the Museum of Comparative Zoology at Harvard) convince us 
that Aepyornis was not as much larger than an ostrich as the egg would lead us to 
believe. The largest Aepyornis was about as tall as the largest moa. 

Archaeopteryx and Archaeornis lived at a time not far removed from that at 
which their stock diverged from some reptilian ancestor. But what was the ancestor, 
and how was the separation accomplished? No positive answers to these questions 
can as yet be given. We know so little that the most that can be done is to formulate 
theories and assemble facts to support or contradict them. Since the origin of flight 
is intimately connected with the origin of birds, that topic may be discussed first. 
Two theories purporting to account for the use of the anterior limbs in flying are 
current. Neither, it must be confessed, is really plausible or particularly well supported 
by evidence drawn from the anatomical characteristics of birds. 

The first is the "cursorial theory," championed by the late Baron Nopcsa, an 


eminent Hungarian paleontologist. He pointed out that the hind limbs of birds are 
not so modified as to assist in flying but are essentially similar to those of bipedal 
dinosaurs. He thought that the first step toward the acquisition of the power of 
flight was the assumption of a semierect posture, in which attitude the hind legs 
were the sole organs of locomotion. On this hypothesis the ancestor was a slender, 
hollow-boned, bipedal creature which ran as readily as it walked. Everyone has seen 
athletes wave their arms as they run, seemingly trying, by beating the air, to pull 
themselves along a bit faster. Nopcsa maintained that the ancestral reptile-bird did 
the same and gained thereby a certain increment of speed, for the scales on the arms 
provided a surface sufficient to be of some use when pressed against the air. According 
to this theory, continued use of the arms as propellers brought constant increase in 
the length and breadth of the scales on them. As the scales lengthened, the pressure 
against the air caused their edges to become frayed, and so scales gradually changed 
to feathers. The broader the arm became, the greater its contribution to the speed 
of the animal. From generation to generation the wings increased in size, till finally, 
on a happy day, becoming at last more important than the legs, they lifted the body 
off the ground, and flight had been attained! 

This theory seems a trifle fanciful, and verges upon the absurd in its assumption 
that feathers evolved first on the wings and then spread all over the body. The rather 
naive idea that feathers are frayed-out scales should not be charged to the discredit 
of Baron Nopcsa. The same opinion has been held by many zoologists and most 
paleontologists, and the statement that feathers are modified scales appears in most 
textbooks. It is only recently that it has been shown that they are fundamentally dif- 
ferent structures, arising from different layers of cells in the skin. Feathers are as 
absolutely confined to the birds as hair is to the mammals. The cursorial theory of 
the origin of flight cannot be taken seriously enough to require detailed refutation. It 
should, however, be noted that there is nothing about the structure of Archaeopteryx 
or Archaeornis to suggest that they were rapid runners. The small pelvis and rela- 
tively feeble hind limbs are strikingly different from those of the cursorial birds of 
the Tertiary and Pleistocene, or even the bipedal dinosaurs. 

The second suggestion is that of William Beebe and W. K. Gregory, who pro- 
posed what is known as the tetrapteryx theory. They suppose that the ancestors of 
the birds were active arboreal reptiles which were in the habit of jumping from branch 
to branch and so from tree to tree. Those individuals with the largest scales on arms 
and legs were able to make the longest swooping jumps; so from generation to 
generation they became better and better gliders. Beating the air with fore and hind 
limbs extended the distance coverable in such leaps, developing wings on all legs, 
the scales being transformed into feathers by the fraying of the edges. After acquiring 
considerable skill in flight by the use of four wings (the tetrapteryx stage), the birds 
discovered that they could do better by using the anterior pair only. Since the posterior 


wings became inactive, they gradually lost their large feathers and reached their 
present condition. The tetrapteryx stage served merely to tide the animals over 
that difficult period during which they were learning to fly. 

The fundamental idea in this theory that is, that the ancestors of the birds 
were scansorial,- arboreal reptiles is good, but there seems to be no evidence that 
birds ever had four wings. It is true that the hind limbs of Archaeornis appear to 
havfc borne a series of long quill feathers, but that they formed a wing is not so certain. 
It has also been asserted that the hind legs of the embryos of certain modern birds 
show lines of bristles which are vestiges of the wing feathers of the ancestor, but 
Heilmann has disproved this statement. 

The tetrapteryx theory, with the posterior wings left out, a sort of Hamlet with- 
out Hamlet, is supported by Heilmann, the Danish ornithologist who has made 
the most important contributions to the solution of the problem of the origin of 
birds. His "gliding theory" includes the fundamental ideas of Beebe and Gregory 
with additions of his own, as will be indicated in later paragraphs. 

As for the question of the ancestry of birds, three theories have been advanced. 
The first and most natural of these is that flying birds evolved from the flying rep- 
tiles, the pterosaurs. The similarities between the two are numerous. Both, in the 
early part of their history, have conical teeth, which in later ages are lost and replaced 
by a bill. In both the skull is held nearly at right angles to the backbone, and the 
brains are much alike. Both have large orbits for the eyes, a ring of sclerotic plates, 
hollow, air-filled bones with dense walls, and a large sternum, an element rarely ossified 
in reptiles. Yet despite the numerous similarities it is impossible that pterosaurs 
should have given rise to birds, for their chief specialization was the enormous de- 
velopment of the fourth finger, a digit of which no bird, even the most ancient, retains 
the least vestige. 

A second theory, with less basis but, until recently, widely accepted, is that 
birds descended from bipedal dinosaurs. The chief features common to the two 
groups are hollow bones in theropodous dinosaurs and birds; posterior branches of 
the pubes in ornithischian dinosaurs, comparable in position to the pubes of birds; 
and ankle joints between the two rows of tarsals. The hind feet of many theropods 
have only four clawed toes, the fifth absent, and the first rotated behind the meta- 
tarsals, as in birds. Finally, the pelvis of some dinosaurs is large and long, with as 
many as eight vertebrae in the sacrum. 

Although such a thought seems not to have occurred to many of the geologists 
and paleontologists who accepted this theory while it was popular, it would be im- 
possible for birds to have arisen from the large heavy-boned ornithischians, in which 
group alone is to be found the so-called birdlike pelvis. Moreover, since birds have 
no prepubis, the homology between the ornithischian and avian pelvis is far from 
being exact. The only possible ancestors are the slender, hollow-boned theropodous 


coelurosaurs, which as Heilmann has shown in his book, The Origin of Birds, have 
many birdlikc characteristics: "Hollow bones of very light structure, exceedingly 
long hind limbs with strong, elongate metatarsals and a 'hind toe,' a long narrow 
hand, a long tail and a long neck, large orbits and ventral ribs these are bird- 
features immediately conspicuous." 

As will be remembered, the coelurosaur lineage persisted from late Triassic to 
late Cretaceous times, which makes it possible to observe some of the evolutionary 
trends in the group even if individual lineages cannot be followed. These trends 
are such as to cause the coelurosaurs to become more and more birdlike with the 
passage of time. Triassic representatives of the group have conical teeth in sockets; 
Jurassic Compsognathus and Ornitholestes have similar but smaller teeth and fewer 
of them; Cretaceous Struthiomimus has lost all teeth and acquired a bill. Similarly, 
the bones of the skull became more slender, and the preorbital openings larger and 
larger, more like those of Archaeornis. The pelvis increased in length, the number 
of sacrals changing from three to five. Perhaps the most important feature of the 
evolution of the group is the fact that in the course of time the fore limbs, short in 
the Triassic Podofesaurus and Procompsognathus, became longer and longer; Heil- 
mann was the first to point out this change, which is opposite to that characteristic of 
the other theropodous dinosaurs. As the arms became longer, the outer fingers were 
lost, until only three remain. The late coelurosaurs were, except for the absence of 
feathers, exceedingly birdlike. The obvious difference in skeletal structure is that in 
them the pubic bones extend downward and forward, instead of backward. But this 
is no insuperable obstacle to the derivation of the birds from them, for the pelvis 
of embryos of birds is of the typical reptilian triradiate type. The distal end of the 
pubis, it is true, is recurved, but the proximal portion projects downward and some- 
what forward (compare Fig. 93). This is interpreted as indicating that the ancestor 
of the birds had a pelvis more like that of theropod than like that of ornithischian 

If one could stop here, it would seem logical to conclude that birds were derived 
from coelurosaurs. But there is a fly in the amber, a tiny flaw in the argument. 
Coelurosaurs had no clavicles. Pterosaurs managed to fly without the aid of these 
bones, but apparently birds cannot. A wishbone, or at least the elements thereof, is 
present in all birds, even the oldest. It is evidently a sine qua non of bird organization. 
No reptile which had lost it could have been ancestral to the group. The present 
opinion is that most of the birdlike features of the dinosaurs are due rather to the 
semierect position in which the animals walk than to the descent of one group from 
the other. Nevertheless, it is generally held that birds and dinosaurs had common 
ancestors, a position ably supported by Heilmann. After a detailed study of the 
osteology, anatomy, and embryology of birds, he has shown many reasons for believing 
that the pseudosuchian reptiles of the Triassic of South Africa, now generally accepted 

FIG. 93. One side of the pelvis of an ostrich (A), as compared with that 
of Archaeopteryx (B). Ilium above, ischium beneath, with backward pro- 
jecting pubis below it. Redrawn after G. Heilmann. At right, the pelvis of 
Eupar1(cria t to show the downward curvature of the pubic bones. One-half 
natural size. From R. Broom. 

FIG. 94. Dorsal and lateral views of the skull of Eupar?(eria, a representative 
of the group (Pseudosuchia) from which both the dinosaurs and the birds 
are supposed to have been derived. Lettering as in Fig. 64. One-half natural 
size. From R. Broom. 

FIG. 94A. A "quarrel scene" from the Upper Jurassic, with Rhamphorhyncus 
taking the part of Mercutio, and Archaeornis that of Tybalt. From a lithograph 
by Charles R. Knight, with his permission. 


as the ancestors of the theropodous dinosaurs, were also those of birds. Unfortunately, 
his argument depends largely upon the interpretation of the elements of the skull 
of Archaeornis, a portion of the skeleton so badly crushed that more than one 
reconstruction is possible. 

The movable quadrate and the lateral bar of the skull are so characteristic of 
birds that one might expect them to be present in the ancestor. Among reptiles, the 
free quadrate is found in the Squamata only, but one seeks in vain among the lizards, 
the most primitive of the Squamata, for any representative suggestive of a bird. They 
have no lower temporal arcade and hence no structure comparable to the lateral bar 
of the birds. Consequently, we are forced to agree with Heilmann that the movable 
quadrate is not a feature of prime importance but was formed independently in 
squamates and birds. Poorly preserved as the skull of Archaeornis is, it does seem to 
show that the quadrate was immovably articulated with the squamosal. Furthermore, 
the marginal bones were not of avian structure, for there was unquestionably an 
ascending process of the jugal behind the orbit, and the quadratojugal must have been 
short. Thus interpreted, the skull of Archaeornis is similar to that of the pseudo- 
suchian, Euparfyeria (Fig. 94), as described by Broom from specimens collected from 
the Triassic of South Africa. The chief difference is that Euparferia, being a diapsi- 
dan, has a dorsal temporal opening, which is absent from Archaeornis and all later 
birds. The parietals, postorbitals, and squamosals of the latter appear to have grown 
together so as to obliterate the upper fenestrae. So far as the skull is concerned, all 
that can now be said is that the pseudosuchian presents a possible model by which 
the skull of Archaeornis may be reconstructed. 

The pseudosuchians were bipedal reptiles with five toes on both front and 
hind feet. The arms were shorter than the legs but well developed. The body was 
covered with scales, and in Euparferia, Broom says, "all the best preserved scales 
were about twice as long as broad and have the long axis lying anteroposteriorly . . . 
from this axis there are distinctly traceable ribs running sideways, in form almost 
representing a feather; we neefd merely imagine the ribs continued beyond the border 
of the scale." The obvious suggestion is that the scale was in process of transforma- 
tion into a feather. One should always be cautious, lest one prove too much. 

Two long bones, a part of the hyoid arch of Eupar\eria, have led Broom to the 
conclusion that this animal had a birdlike tongue. Other avian characteristics men- 
tioned by Broom or Heilmann are: cervical and presacral vertebrae of approximately 
the same number; ribs double-headed, with small uncinate processes; ventral ribs 
present; clavicles long and slender, forming part of a shoulder girdle which has all 
the elements of that of a bird. The pelvis of the pseudosuchian is particularly im- 
portant, for, although it is attached to only two sacral vertebrae, it is like that of the 
embryo of a modern bird, the pubis projecting forward in its inner portion, down- 
ward and slightly backward in the outer. According to Broom, the tarsus foreshadows 


that of the birds; the metatarsals are elongate, the third best developed, an avian 
rather than a reptilian characteristic. 

Heilmann has given a picture of a hypothetical animal intermediate in charac- 
teristics between the pseudosuchians and the Upper Jurassic birds. His Proavis, "no 
longer a reptile and not yet a bird," is described as an inhabitant of the trees of the 
Triassic forest, where it hopped from branch to branch and gained distance by gliding 
with outstretched arms. The exigencies of arboreal life led to the perfection of a 
grasping type of foot and to the elongation of the arms. The long slender body and 
long tail are described as covered with long, broad scales, which were beginning to 
change into feathers, particularly on the arms, then in the initial stages of trans- 
formation into wings. The stimulus derived from the pressure of the air on the 
tips of the scales is supposed to have caused their further growth, particularly where 
the pressure was strongest, as on the posterior margins of the limbs and along the 
sides of the tail. Thus eventually an Archaeopteryx stage was reached. About the 
skull one can speak with less assurance, but it appears to Heilmann to have been 
large, with numerous conical teeth, its structure much more reptilian than avian. 

Proavis may or may not have existed. It is one of the "missing links"; toward 
its recovery, exploration should be directed. If the pseudosuchians were really the 
ancestors of the birds, one would expect to find the remains of the annectent group in 
the older Jurassic strata of Africa. For the present, however, it will be well to keep 
an open mind as to avian ancestry, for although Heilmann has presented cogent 
arguments for the acceptance of his theory, they are based largely upon the inter- 
pretation of the single badly preserved skull of Archaeornis. 

FIG. 95. Restoration of Archaeopteryx, redrawn after W. P. Pycraft. 


He that useth many words for the explaining of any subject, doth, like the cuttle-fish, 
hide himself for the most part in his own ink. 

John Ray, On Creation 

Some animals seem predestined to obscurity. Who has dramatized the life of a 
bryozoan or written odes to a brachiopod? Other creatures, because of size, color, 
repulsiveness, or beauty, gain a place in the popular imagination, and become im- 
mortalized in literature or art. The lowly scallop was carried to honor as the badge 
of the Pilgrims; the shell of the triton appeared as a decorative motif in Grecian 
times. No marine invertebrates, however, have been more fully adopted by artists 
and writers than the cephalopods. Where is he who has not shuddered with Hugo 
as Gilliatt fought the octopus, and whose adolescent ears have not been assailed by 
recitations of that quintessence of Victorianism with the ringing moral: "Build thee 
more stately mansions, O my soul"? The squid, octopus, cuttlefish, and pearly 
Nautilus have characteristics of one sort or another which have made them well 
known. Some are swift, dramatically ejecting ink to baffle pursuers; a few are giants, 
hasty glimpses of which probably inspire some of the sea serpent stories; others are 
of devilish repute because of their horrid appeal to the imagination. The pearly shell 
of the Nautilus has for centuries made it valuable, a thing to be encased in gold, 
an ornament to the table of the highborn and wealthy. 

Although they are so well known, the cephalopods are at present a relatively 
small group. They fall readily into two subclasses. The more abundant, the squids, 
cuttlefish, and octopi, lack an external shell, so that they might be called the naked 
cephalopods, although, because they have only two gills, their scientific name is 
Dibranchiata. The others, represented at present only by three species of Nautilus 
in the Indian Ocean and the parts of the western Pacific adjacent to southern Asia 
and the Philippine Islands, are sometimes denominated the shelled cephalopods, or, 
technically, the Tetrabranchiata, since they have four gills. 

Examination of a squid, cuttlefish, or octopus will show that all have more or 
less elongate or bulbous saclike bodies, set off from the head by a constricted region 
resembling a neck. The head of the squid has ten, that of the octopus eight, 
more or less elongated processes, the arms or tentacles. These are supposed to repre- 
sent the foot; hence the name Cephalopoda, "head-footed." The tentacles are pro- 
vided with suckers or hooks, effective organs for grasping and holding. Many 


dibranchiates have two especially elongated ones, modified to serve in transferring 
the male sperm to the females; these arms, in the largest of living squids, animals 
with bodies five feet in circumference and fifteen feet long, reach a length of some 
thirty-five feet. Sixty years ago tales of such creatures were classed as myths and 
sailors' yarns, and depicted in various paintings and engravings showing the attack 
and overthrow of skiffs and small sailing vessels by the "octopus." Now it is known 
that the tall stories have a basis in fact. 

The eyes of the naked cephalopods are marvels in the invertebrate realm. Pro- 
vided with a lens, cornea, and lid folds, they approach in perfection those of the 
vertebrates. It is doubtless the possession of this type of eye which has gained for 
the octopus the reputation of being the most intelligent of invertebrates. Lying con- 
cealed among the shadows of the reefs which form a congenial habitat, this keen- 
sighted beast has been observed to dart out only when there is obvious opportunity 
to capture its favorite prey. 

Naked though such cephalopods are, most are not entirely shell-less. If a squid's 
back be cut open, a thin, flexible pen is discovered within the folds of the mantle 
which envelops the body (Fig. 96, at left). Too weak to be of any use as a support, 
it is obviously a vestigial structure. The octopus has no internal shell, but within the 
back of the cuttlefish is found the broad, porous, calcareous plate familiar to bird 
fanciers as the cuttlebone. Another modern dibranchiate has still more of a skeleton. 
Thousands of small, loosely coiled, chambered shells appear at times on the shores 
of certain islands in the tropical and subtropical oceans. They somewhat resemble 
Nautilus, except for their small size and the fact that the whorls are not in contact. 
Although long known by the name of Spirula, the living animal has been found 
only recently. This shell is not an external one, but is held within the mantle, near 
the posterior end of the body. 

Without fossils it would be impossible to understand the significance of the 
curious hidden shells of the modern naked cephalopods, but instead of trying to 
trace their ancestry in detail it is easier to step back at once to the earlier deposits 
of the Mesozoic, there to examine some relatively ancient remains. 

Residents of England, France, and Germany have for centuries been familiar 
with cigar-shaped stones commonly found in clay pits. Popularly known as "thunder- 
bolts," they were long supposed to be of extraterrestrial origin, so it is not surprising 
that even as late as 1750 they formed a part of the current pharmacopoeia, being ad- 
ministered in powdered form as a medicine where a deficiency in calcium was in- 
dicated. The oldest "thunderbolts/* or belemnites (Fig. 96), are found in strata of 
Triassic age, but they are much more common in Jurassic and Cretaceous deposits. 
Some are small; others are as much as eighteen inches long and more than an inch 
in diameter. The ordinary specimen is cylindrico-conical, pointed at the posterior 
end, deeply pierced by a conical cavity at the other. Except for this the body is solid. 

FIG. 96. At left, a modern squid with its pen. At right, various belemnites. 
The tall figures are restorations, showing the three parts of the shell; the short 
ones actual specimens. All from F. J. Pictet, Traitt dc palcontologie. 

FIG. 97. Three theoretical stages in the evolution of the belemnites. Upper 
figure, animal within shell. Middle figure, animal outgrows and engulfs 
shell; part of the living chamber is resorbed. Lowest figure, all the living 
chamber except the dorsal proostracum is lost, and a guard covers the phrag- 
rnocone. Original drawings by Bradford Willard. 

FIG. 98. Various types of ammonoid sutures. Above, at left, a goniatite 
with simple angular inflections; center, two views of an ammonite with the 
ceratitan suture, with denticulate lobes and smooth saddles; and at right, a 
fragment of an ammonite with a highly complicated suture. All from F. J. 
Pictet, Traite de paleontologte. Below, a drawing of the most complicated 
suture known, that of the Triassic Pinacoceras. From F. von Hauer. 

FIG. 99. A complete ammonite shell, retaining the living chamber and 
lateral lappets. Note the position of the hyponome below the lappets and the 
notches for the eyes above them. The position of the shell is, of course, un- 
natural. From F. J. Pictet, Traitt de paleontologie. 


Broken transversely, it is seen to be made up of radially arranged prisms of calcite; 
when cut longitudinally it may be noted that the prisms are crossed by lines cor- 
responding with the outline of the exterior. In other words, the belemnite was built 
up by the addition of successive external layers. Occasionally specimens are found 
with a conical shell in the cavity. This shell is subdivided by a set of curved partitions 
just as Nautilus or Spirula is. Each partition has a small circular opening on the 
ventral side, a characteristic which proves that the belemnite is really a cephalopod. 
At the apex is a minute globular chamber, the significance of which will be mentioned 
later. Still more rarely, even more complete specimens are found. In addition to 
the "thunderbolt" and the conical chambered shell imbedded in it, these have a broad 
thin extension of shelly matter which is an anterior, dorsal extension of the upper 
part of the latter. An entire belemnite consists, therefore, of three parts: a solid por- 
tion or guard, a chambered shell known as the phragmocone, and a thin flattened 
dorsal extension of the latter, the proostracum. 

It has already been pointed out that the guard was built up by the addition of 
successive external layers. This shows that it was formed later in life than the phrag- 
mocone, being deposited upon its surface. It also indicates that the shell of the 
belemnite was internal. This evidence is corroborated by the fact that many guards 
show on their surfaces ramifying canals which are impressions of vessels carrying 
the circulatory fluids. The belemnites are, therefore, skeletons of naked cephalopods. 
Fortunately, a few specimens retain impressions of tentacles. None is absolutely 
complete; hence there is some difference of opinion about a reconstruction. Many 
investigators have held that there were ten arms, as in modern squids; a careful 
student, Professor Abel of Vienna, finds evidence of only six. They appear not to 
have been provided with suckers but with hooks, equally useful but less specialized 
organs of prehension. 

Since the belemnites are the oldest naked cephalopods, it is natural that paleon- 
tologists should think of them as the ancestors of the modern ones. Apparently the 
internal shell has been progressively reduced (Fig. 97). The modern cuttlefish seems 
to retain only the modified proostracum, the incurved tip of which may represent 
a vestige of the phragmocone. The pen of the squid is regarded as but a vestige of 
the proostracum; the octopus has lost even that remnant of shell. The Upper Jurassic 
strata at Solenhofen have afforded specimens of typical pens of squids, and Jurassic 
strata in England and Germany have produced excellent cuttlebones, some of them 
with sacs of fossil ink, a substance so like the modern sepia that, properly ground, 
it has been used by draftsmen in depicting the animal which furnished the pigment. 

So much regarding the naked cephalopods. As far as they are now known, they 
appeared first, with heavy internal shells of threefold nature, in the Triassic, were 
extremely abundant in this expression throughout the Mesozoic, but lost a great 
part of the skeleton during the Tertiary. At the present time they have reached their 


culmination, in size at least, in the gigantic squids and octopi of Atlantic waters. 

If the modern squid and octopus may be compared with those originally of good 
family, somewhat decayed, but capable of rising again to prominence by taking 
advantage of recent happenings, Nautilus is the peer without land or income, with 
a wonderful ancestry but no present position. So poor, indeed, is Nautilus that al- 
though we know his lineage for hundreds of millions of years we do not know the 
full life history of a single individual. 

For the moment, however, we are interested chiefly in the adult, which is well 
known. The shell, like that of a snail, is coiled, but in one plane, after the fashion of a 
watch spring, rather than in a screwlike spiral. Unlike that of a watch spring, how- 
ever, the last complete turn or volution envelops the others, almost completely hiding 
them. When the shell is cut along the median plane, the section reveals several volu- 
tions. About halfway back in the outermost there is a concave partition or septum 
which forms the posterior wall of the living chamber, the habitation of the animal. 
Numerous septa, approximately equally spaced, subdivide the cavity behind the 
living chamber into a series of empty compartments. Each septum is perforated below 
the center by a small opening, around the margin of which the shell is prolonged a 
short distance backward. The tube extending through these openings to the apical 
chamber is the siphon, and the shelly sheaths are the siphonal collars or funnels. 
When Nautilus emerged from the egg it formed a thimble-shaped shell about itself. 
As it grew, it added calcareous material at the anterior margin. When it reached a 
certain size, the animal for some unknown reason ceased for a time to increase the 
length of its shell, pulled the posterior end of its sac-shaped body forward, and secreted 
a calcareous partition behind it. The contour of this septum corresponds to the shape 
of the posterior end of the body. Growth was then renewed, and the process repeated 
until adult size was reached and all the septa were formed. Unfortunately this activity 
is known only by inference from the fact, recently observed by Dr. Rudolf Ruede- 
mann, that there is a coincidence between the number of septa and the number of 
resting stages which can be identified by a study of the lines of growth on the shell. 
It is probable that the so-called resting stages represent periods when food was scarce 
and that the pulling forward of the posterior end of the body was due to shrinkage 
during a time of underfeeding. One of the most interesting experiments which could 
be performed would be to raise young Nautili, but although repeated attempts have 
been made no one has yet been able to obtain fertilized eggs or larval specimens. 
Professor Arthur Willey, who devoted a year and a half of his life to this problem, 
finally succeeded in keeping the animals alive in natural aquaria in the Philippines 
and actually got eggs, but no fertile ones. Nautilus presents a profitable subject of 
research for a zoologist with plenty of time, money, and ingenuity. 

Since the septa are deposited later than that part of the shell adjacent to which 
they lie, it is obvious that they are wholly internal to it. There is nothing whatever 


on the exterior which indicates the location of one. Each meets the interior of the 
shell along a line which corresponds with the shape of the partition. If the septum 
were simply concave, like a watch crystal, and the shell conical, then the line along 
which it met the shell would be a circle. As a matter of fact, the septum in modern 
Nautilus does not have this simple shape; hence the line, which is called the suture, 
turns forward and backward in a definite way. If the interior of a Nautilus were com- 
pletely filled with plaster of Paris and the shell ground off, the edges of the septa, 
the sutures, would become visible, and it would be seen that they make sweeping 
curves backward and forward. Where such a curve is convex backward toward the 
apex, the suture is said to have a lobe; where it is convex forward toward the 
aperture, there is a saddle. Lobes and saddles alternate with one another all around 
the edge of a septum. When cephalopods are studied, it is necessary to have some 
specimens from which the shell has been naturally or artificially removed. 

Naked cephalopods have had a relatively short history, but shelled forms, more 
or less like Nautilus, have existed since the Upper Cambrian perhaps since the 
Lower Cambrian From early Ordovician days till the end of the Mesozoic they 
were abundant. At the end of the Cretaceous, however, some great calamity befell 
them, and only a few managed to survive through the Tertiary to the present. Little 
or nothing is known concerning the soft parts of these ancient animals, but the shells 
are so similar to those of living examples that all are grouped together as the great 
subclass Tetrabranchiata. This should not be taken to mean that paleontologists 
assert that the extinct representatives of the group had four gills. There is reason 
to believe that they had more than two pairs. 

It is readily seen that the shell of Nautilus is, in effect, a coiled, chambered tube. 
If one could be straightened out and restored to a really conical form, it would re- 
semble the sort of cephalopod which was common throughout most of the Paleozoic. 
Many of the older representatives of the group secreted long slender conical shells 
with simple septa and sutures. Such are called orthoceratites, or orthoceracones, be- 
cause of their resemblance to straight horns. They reached their maximum in size 
and variety during the Ordovician and the Silurian, but are found in strata as young 
as the Triassic. A few of the Ordovician specimens are fifteen feet long, and some 
Silurian individuals had a length of seven or eight feet. Although they did not reach 
the gigantic size attained by some modern squids, they were large for their times. 

The oldest cephalopods obviously related to Nautilus are a few found in Upper 
Cambrian rocks. In the Lower Cambrian, however, there are two small forms, one 
of which, Salterella> is common and widely dispersed, whereas the other, V olborthella, 
is found chiefly in countries adjacent to the Baltic. Both are represented by small 
straight or somewhat curved tubes which consist of a series of funnels nested together. 
They are generally considered to be the oldest cephalopods, but show some peculiari- 
ties which suggest that they may have other relationships. 


By no means all Paleozoic cephalopods which resemble Nautilus are ortho- 
ceracones. Some have curved shells, in which case they are called cyrtoceracones. A 
few have one complete volution, although the whorls are not in contact, and are 
known as gyroceracones, a term sometimes abbreviated to gyrocones. Still others 
are fully coiled, but in such a way that the whorls are barely in contact, so that all 
can be seen in lateral view. This is a planospiral coiling which produces the evolute 
nautilicone. As a modification of this basic pattern there are some which in 
the earlier whorls are planospiral but the later-formed portion of the shell is a 
straight tube. Such are known as lituiticones. Some late Paleozoic nautiloids are 
as tightly coiled as modern Nautilus, the last volution embracing and covering 
the earlier whorls, forming a fully involute nautilicone. Needless to say, there are 
all gradations between evolute and involute shells. Finally, there are a few ancient 
relatives of Nautilus which resemble gastropods in having an asymmetrical spiral 
shell, a torticone. 

It should be borne in mind that this is merely a rough and unscientific grouping 
by shell forms, which indicates nothing about relationships. Two orthoceracones may 
be closely or distantly related. They may belong to the same species or to different 
suborders. There are families in which nearly all forms are represented. 

All Cambrian. Ordovician, and almost all Silurian cephalopods have simple septa 
and sutures. Many Devonian and later specimens, however, have more complicated 
ones. Some have angular, pointed lobes, or even angular saddles. Others have more 
elaborate sutures, the lobes and saddles being subdivided by secondary inflec- 
tions. Such forms became increasingly common during the late Paleozoic and 
were the dominant Mesozoic shelled cephalopods. This obvious difference from 
Nautilus led to the investigation of other characteristics and to the consequent 
subdivision of the Tetrabranchiata into two great orders, the Nautiloidea and 
the Ammonoidea. 

Most ammonoids (Fig. 98) can readily be distinguished from nautiloids by the 
greater complexity of the sutures, but since there are exceptions to this rule other 
criteria have been sought. One of the differences lies in the position of the siphonal 
tube. The funnels of Nautilus are short, extending only a millimeter or two back of 
each septum. Most fossil nautiloids and ammonoids, however, have a calcareous 
tube which is continuous through all the deserted chambers. In some it consists 
of elongated siphonal collars which extend from one septum to that behind; in 
others the collar is short, but a calcareous tube of another origin bridges the gap to 
the next partition. Such a continuous tube, however formed, is termed a siphuncle. 
In all ammonoids the siphuncle is in contact with the inside of the shell, and in 
every family but one it is ventral in position; that is, it lies just within the outer side 
of the whorl, if the shell be a coiled one. In the case of the one exception, it is on the 
dorsal side. In most nautiloids, on the other hand, the siphuncle is not in contact 


with the shell; it may be central, or above or below the center. In cases where it is 
actually ventral it is large, whereas in all ammonoids it is small. 


















FIG. 100. Diagram to illustrate the approximate parallelism between shell 
forms in the ammonoid and nautiloid cephalopods. 

Another and more important difference is one that can be determined only if 
the apex of the specimen be visible. The initial shell of all ammonoids is a minute 
spherical test which remains throughout life attached to the apex of the cone. In 
most coiled shells it is completely buried by the later whorls but can be found by 


breaking back to the center. Nautiloids lack such an initial chamber. At the apex 
of a Nautilus is a scar which has been interpreted as indicating that the initial shell 
was membranous or chitinous, not capable of preservation. Similar scars are present 
on fossil nautiloids. None has been found with a spherical calcareous nucleus. This 
last statement has been disputed on the evidence of certain straight shells found in 
the Devonian of New York, but they were probably not correctly identified. 

The oldest ammonoids, which are found in the Silurian, and most of the late 
Paleozoic ones, have relatively simple sutures. The majority have a few rounded 
or angular lobes and saddles, and a narrow V-shaped ventral lobe. The angularity 
of some of the lobes or saddles suggested the common name, goniatites, but the word 
is not used in modern classifications. Most goniatites (Fig. 98, at left) are small, 
although a few are as much as ten or twelve inches in diameter. Since they abound 
in certain regions, particularly in Upper Devonian and Upper Carboniferous strata, 
they have been of great use in stratigraphy and correlation. 

There is a gradual transition from goniatites to true ammonites. The latter differ 
only in having secondary inflections of the suture (Fig. 98). Consequently, the appear- 
ance of a single saddle within a lobe, or a lobe within a saddle, would signalize the 
transition from one group to the other. Various goniatites gave rise to diverse stocks 
of ammonites, a progress accomplished during the Carboniferous and the Permian. 
Like their progenitors, the Paleozoic ammonites were all of small size, but during 
Triassic times conditions appear to have been extremely favorable to members of 
this group, for then they increased tremendously in abundance, size, and particularly 
in complexity of sutures. In fact, one of the Triassic forms, Pinacoceras (Fig. 98, 
below), had the most complicated sutures, which means, of course, the most com- 
plexly curved septa, of all known cephalopods. 

For some as yet unexplained reason the end of the Triassic was a critical era for 
the ammonites. According to specialists, representatives of all but one of the many 
families then in existence perished at that time. Members of this one family, fortu- 
nately, weathered their adversities. Like Noah's children they increased, multiplied, 
and replenished the earth, or, rather, the sea. Their evolution must have been ex- 
traordinarily rapid, for the oldest Jurassic rocks, the Liassic, are the greatest of the 
world's repositories of the remains of these creatures. I have stood on strata of this 
age at Lyme Regis, in Dorsetshire, and from one spot counted over two hundred 
ammonites on the surface of a single layer, each of them more than a foot in diameter. 
Although the Triassic was the time of culmination of the ammonites in variety, the 
Upper Jurassic saw their giants, some of them five, a few ten feet in diameter. The 
group remained a dominant one in the seas till the end of the Cretaceous. Then, 
suddenly, they were all gone. Why? No one really knows. 

Although much has been written about the overspecialization and degeneracy of 
the Cretaceous ammonites, their peculiar forms have lately been interpreted as due to 

FIG. 101. Chart to show one interpretation of the history of the cephalopods. 
Solid lines connect members of the Nautiloidea, dashed lines the Dibranchiata, 
and dots and dashes the Ammonoidea. All figures conventionalized. 



their habits. As to overspecialization, we have nothing but surmises, for we do not 
know to what conditions they were adapted. Mere complexity of suture cannot have 
been an important factor, for that had reached its climax during the Triassic. More- 
over, many of the Cretaceous ammonites had relatively simple sutures; yet they 
succumbed with the rest. 

Geologists have found the ammonites to be the most important of all Mesozoic 
fossils for purposes of identifying and correlating strata. Evolution was rapid in the 
group, particularly during the Triassic and Jurassic. Careful students have noted 
changes in ammonite faunas and progress in ammonite evolution from bed to bed 
in the vertical succession, a circumstance that makes possible detailed correlations 
within restricted regions. On the other hand, it has been found that some ammonites 
were distributed rapidly over extensive areas, probably through transportation by 
oceanic currents, either as members of the true plankton or possibly as tenantless, 
gas-filled shells. This makes it possible to correlate strata over wide areas, as, for 
instance, the Triassic of southeastern California and Nevada and that of the countries 
bordering on the Mediterranean. 

The ammonoids have a series of shell forms (Fig. 100) much like that of the 
nautiloids, although an even longer series of names has been proposed. A few of the 
earlier goniatites are as straight as the orthoceracones. They are called bactriticones. 
A few are loosely coiled, at least in young stages. These are the mimoceracones, which 
correspond with the gyroceracones. Many, the ophiocones, have evolute coiling, 
whereas others, the ammoniticones, are involute; as among the nautiloids, there are 
all gradations between the former and the latter. Among the so-called degenerate 
forms there is great variety, although only a few types have been named. Ancylo- 
ceracones have the earlier whorls evolute, the larger part of the shell straight, curved, 
or hooked. Crioceracones are much like gyroceracones, the whorls not in contact. 
Baculiticones are apparently straight shells, but complete specimens have at the apex 
a minute involute coil. The coil is so seldom seen, however, that such are commonly 
called "straight ammonites." Turreted snaillike shells are much more common than 
among the nautiloids. They occur in strata as old as the Triassic and are known in 
general as torticones, more particularly as turriliticones, if coiled with the whorls 
in contact. Some shells, such as the Japanese Nipponites, with a shape that has been 
compared to that of a model of the path which a particle follows during an earth- 
quakej or the western American Emperoceras, which in its young stages was an 
ammoniticone, later became an ancyloceracone, and in the adult was a torticone, 
practically defy classification. They may be called hysterogenicones. 

It is of some interest to note that the Nautiloidea are the ancient and central 
stock of the Cephalopoda, that they reached their climax in the Mid-Paleozoic, and 
that they persist to the present practically unchanged from the state which they had 
reached in the late Paleozoic. From the Nautiloidea sprang the Ammonoidea, prob- 


ably during the Silurian time. This group had a long period of luxuriance in the 
Mesozoic but disappeared abruptly. There is difference of opinion about the origin 
of the Dibranchiata, but the spherical initial chamber of the phragmocone seems to 
indicate that they are descended from the ammonoids. Late in their arrival, they 
flourished throughout the Mesozoic, became greatly modified during the Tertiary, 
and eem to be enjoying a second period of vitality today. 


The bee is enclosed, and shines preserved, in a tear of the sisters of Phaeton, so that it 
seems enshrined in its own nectar. Martial, c. A.D. 90 

Whence we see spiders, flies, or ants entombed and preserved forever in amber, a more 
than royal tomb. Francis Bacon, 1623 

I saw a flie within a beade 
Of amber cleanly buried. 
t Robert Herrick, c. 1648 

At the meeting of the Royal Society were exhibited some pieces of amber sent by the 
Duke of Brandenburg, in one of which was a spider, in another a gnat, both very entire. 

Evelyn, Diary, March 24, 1682 

Ever since Neolithic times, four thousand or more years ago, Baltic amber has been 
an article of commerce. For at least twenty centuries the animals trapped therein 
have been a focus of interest, since by their singularly perfect preservation they were 
easily recognized. The Hispano-Roman epigrammatist Martial, the politician-philoso- 
pher Bacon, and the clergyman-poet Herrick, whose special claim to fame is that 
he taught his favorite pig to drink from a tankard, had no more difficulty in identify- 
ing these fossils than did the members of the Royal Society. Fossil insects are, there- 
fore, no novelty. They were known hundreds of years before there was a science of 
paleontology. Nevertheless, it was only during the early decades of the twentieth 
century that this branch of the science was put on a satisfactory basis, primarily as 
the result of the researches of Anton Handlirsch, R. J. Tillyard, A. V. Martynov, and 
F. M. Carpenter. The last, a colleague and erstwhile student of mine, furnished 
material for the present chapter and wrote several pages of it. 

It is impossible in a book of this nature to do justice to the insects. Although 
they are only a subordinate branch of the phylum Arthropoda, and although only 
one-third or one-fourth of them have yet been described, the known species outnumber 
all other animals by at least two to one. Several thousand (10,400 in 1930) species 
of extinct insects are now known, but recent discoveries indicate that we are just 
beginning to become acquainted with the fossils. As will be shown in the sequel, 
whole chapters of their evolution are still unknown. 

Insects are commonly thought of as those invertebrates which have the power 
of flight. As a matter of fact, not all members of the class have wings. A better defini- 


tion is suggested by the old name of the group, the Hexapoda. The class may be 
defined as that group of arthropods which is characterized by the possession of three 
pairs of walking legs. The body is divided into three regions, head, thorax, and 
abdomen, separated by more or less pronounced constrictions. The appendages of 
the head include a pair of antennae and the mouth-parts, consisting of a labrum or 
upper lip, a pair of mandibles, one pair of maxillae, and a median labium, or lower 
lip. In the more generalized insects the mouth-parts are mandibulate, that is, adapted 
for chewing; but in many others they are modified for piercing and sucking. Most 
adult individuals possess two sets of eyes, one pair of compound ones and three 
simple eyes or ocelli. The thorax, which bears the organs of locomotion, consists of 
three segments, each with a pair of legs. Wings are borne on the second and third 
thoracic segments of most living members of the subclass Pterygota, although in the 
true flies (Diptera) the hind pair are absent. The segmentation of the abdomen 
of living insects is distinct in most cases. Primitive forms, and the young of others, 
show eleven segments in addition to a terminal process. The abdomens of most lack 
appendages, but some have jointed cerci, which appear to belong to the eleventh 
segment. The males, which have the opening of the seminal ducts on the ninth 
segment, in sonje cases have claspers in that region. The oviducts of the females 
open between the seventh and eighth segments, where some have ovipositors, organs 
that guide the eggs into localities favorable to the hatching and feeding of the embryos. 
The wings are of much importance to the paleoentomologist, and a detailed 
knowledge of their structure is essential, since about half of all the fossils are known 
from these organs only. Fortunately, the "veins" of the wings have been shown to 
be of the greatest value in the identification of the various groups. Although only 
a few primary trunk veins are present, the possible arrangements of longitudinal 
and cross-veins are almost unlimited. Experience with the modern forms, which 
constitute more than 95 per cent of all described species, indicates that the wing 
venation can be relied upon for the separation of genera and species of most orders. 
After a comparative study of the wings of all modern and some fossil insects, Pro- 
fessors J. H. Comstock and J. G. Needham concluded that the principal longitudinal 
veins of the wings could be homologized, and that in general the homologies could 
be determined by the structure of the immature wings. They were thus able to con- 
struct a hypothetical venational pattern representing their conception of the original 
wing structure. The insects which possess wings most closely resembling this hypo- 
thetical type are regarded as the most primitive. It is interesting to note that, although 
the Comstock-Needham interpretation has in general proved satisfactory to the student 
of recent insects and has been widely adopted, paleoentomologists have found it 
necessary to make modifications of the scheme. This is merely one example of the 
necessity of studying the ancient representatives of a group before making generaliza- 
tions. Professor A. Lameere, who has studied fossil insects as well as living forms 


and has a knowledge of the results obtained by Tillyard and others during the first 
two decades of the twentieth century, has formulated a modification of the Comstock- 
Needham interpretation which bids fair to replace it. 

The oldest insects now known are found in strata of the Alleghany series, which 
are, next to the Pottsville, the oldest rocks of the Upper Carboniferous (Pennsylvanian) 
system. Unfortunately, they are full-fledged Pterygota, with no indication of the 
particular group of arthropods from which they sprang. Many of them belong to an 
extinct order, the Palaeodictyoptera, creatures which, although they do not reveal 
their ancestry, are in some respects more simple than any later forms (Fig. 102, at 
left). True insects, with two pairs of well-developed wings (vestiges of a third pair 
are present on the first thoracic segment) and three pairs of legs, these animals 
nevertheless have primitive characteristics, such as the conspicuous segmentation of 
the abdomen, caudal cerci, and a simple venation, with seven principal longitudinal 
veins crossed by numerous transverse ones. Moreover, the wings were broadly joined 
to the thorax, and hence strictly lateral, incapable of being folded back over the body. 
The Carboniferous members of the order had chewing mouth-parts. The group 
is best represented by fossils in Pennsylvanian strata, but a few individuals indicate 
its continuance in Permian times. Among the latter is one species which had the 
mouth-parts modified for sucking purposes. This specialization has led some students 
to place this species in another order, the Protohemiptera. 

The Palaeodictyoptera are not only the oldest insects now known but the probable 
ancestors of all other winged ones. It is not likely, however, that they will con- 
tinue to hold this proud position, for simpler fossils of the class must inevitably be 
found in Mississippian and Devonian rocks. Even now there is evidence that a 
wingless form is present in the Rhynie chert (Mid-Devonian) of Scotland; antennae 
similar to those of modern springtails have been described, but the heads to which 
they are attached are so poorly preserved that many paleontologists have doubted 
the identification. It is understood, however, that bodies have been found which 
show that the animal really belongs in the order Collembola. Unfortunately, de- 
tailed descriptions have not yet been published. If this claim is substantiated, it will 
confirm the inference which entomologists have already drawn, that wingless insects 
(Apterygota) preceded those having the power of flight. 

Although the Palaeodictyoptera were the simplest hexapods existing during 
Upper Carboniferous times, they were not the most abundant. This honor belongs 
to the Blattaria or cockroaches, an order which reached its culmination during that 
period. Particularly well suited for longevity by their omnivorous food-habits and 
their relatively simple structure, they have persisted to the present day. Their decline, 
nevertheless, was a rapid one, for Carpenter has shown that, although they made up 
about 60 per cent of the Upper Carboniferous insectan fauna, they were reduced to 
approximately 35 per cent in the Permian, 7 per cent in the Mesozoic, and only i per 



cent in the Tertiary and Recent. This refers to species, not individuals; the "Croton 
bugs" are still with us. 

Cockroaches and Palaeodictyoptera were the most common Pennsylvanian in- 
sects, but considerable differentiation had been achieved at that time, for thirteen 
orders are represented, all but one (Blattaria) now extinct. Most striking of all were 
the Protodonata, creatures which resembled dragonflies. Most of them were larger 
than existing dragonflies and one of them, Meganeura, an inhabitant of western 
Europe, had wings a foot or more in length. These huge forms were not abnormal 
but, rather, oversized members of a group which, like the dinosaurs a million years 
later, tended to become giants. They spread widely over the northern hemisphere 

FIG. 102. At left, Stenodictya, a palaeodictyopteron from the Upper Car- 
boniferous of France, possessing winglike lobes on the prothorax and a simple 
venation. After Anton Handlirsch. At right, Protelytron, a member of the 
extinct order Protelytroptera, from the Lower Permian of Kansas. The fore 
wings resemble those of beetles. Original drawing by F. M. Carpenter. 

and were abundant during the early Permian. Recently Carpenter found in the 
Kansan Permian a species which was apparently somewhat larger than the Car- 
boniferous one. All the Protodonata were predaceous, feeding on other animals, 
probably mostly on insects, and the large creatures must have consumed an enormous 
number of individuals. It is not improbable that they were in some measure re- 
sponsible for the reduction in the number of roaches in the Permian. Another 
Upper Carboniferous order which deserves special mention is that of the Megasecop- 
tera, a group represented in the Permian by a specialized branch termed the Proto- 
hymenoptera. Most of the other Carboniferous insects belong to orders with 
characteristics more or less like those of the grasshoppers, making up what has been 
called the protorthopteroid complex. Among them may be the ancestors of several 
modern orders which have not as yet been traced back to the Permian. 

The insectan fauna of the Carboniferous is decidedly archaic; that of the suc- 
ceeding period is much more modern, for in Permian strata at one locality or another 


are found representatives of seventeen orders, eleven of which still exist. Only five 
of them originated in Carboniferous times; the others are newcomers. For some 
reason, insectan evolution was rapid at this time. Since hexapods were not then 
specialized for the pollination of plants or for feeding on their nectar (there was 
none), it may be inferred that whatever stimulus led to the change was probably 
geographic rather than organic. Only two probable causes occur to one. The first is 
the general refrigeration which brought on the Permian glaciation; the second, the 
complicated series of changes which must have accompanied the draining of the 
Coal Measure swamps. These two factors may have acted together. Although their 
effect upon the vegetation over most of the northern hemisphere was neither sudden 
nor particularly striking, changes did occur, and they must have had some effect 
upon the evolution of the first aviators. 

FIG. 103. Protentomobrya, a springtail (Collembola) contained in Cre- 
taceous amber from Manitoba. Original drawing by J. W. Folsom. 

Perhaps a few words about some of the more important orders which appeared 
in the Permian may be useful. The geological ranges of insects, as they appear in 
even the most recent textbooks, are so inaccurate that a summary, for which Carpenter 
is responsible, appears to be needed. 

The existing orders which appear for the first time in the Lower Permian are: 
Mecoptera, Neuroptera, Odonata, Psocoptera, Plectoptera, Embiaria, and Homoptera; 
in the Upper Permian, Perlaria, Thysanoptera, and Coleoptera. 

The Mecoptera (scorpion flies) of the Permian are minute, some with a wing 
expanse of only 10 mm. Relatively more numerous than now, they made up about 
9 per cent of the species, whereas their present status is less than .04 per cent. The 
early Neuroptera were in an advanced state of evolution; both the Planipennia (lace- 
wings) and the Megaloptera (alder flies) had been differentiated. The earliest true 
dragonflies (Odonata), unlike the Protodonata, were small creatures, scarcely more 
than an inch and a half across the wings. More vital than their gigantic cousins, the 


real dragonflies have persisted to the present, reaching their culmination in the Upper 
Jurassic, whereas the Protodonata died out in Triassic times. 

The occurrence of true Homoptera in the Lower Permian is of much significance, 
since they appear to have been the most highly specialized members of their class 
at that time. Not only did they possess' mouth-parts adapted for sucking juices from 
plants, but the wing venation was as greatly reduced as that of some existing members 
of the order. The Psocoptera or bark lice were abundant during the Lower Permian; 
over five hundred specimens have been found in the Kansan Permian alone. Unlike 
the modern species of the group, in which the hind wings are much reduced, the 
Permian bark lice had two pairs of similar wings with a venation much like that 
of the Homoptera. The Plectoptera, mayflies, are not uncommon in Lower Permian 
strata. Their nymphs were aquatic, like those of the present. The adult mayflies of 
the time, like the psocids, had two pairs of similar wings, whereas the modern ones 
have greatly reduced hind wings. One of the most surprising facts about these oldest 
known mayflies is that they had the peculiar adult molt, or ecdysis, so characteristic 
of living ones. 

In addition to these modern orders of insects and a few extinct ones which per- 
sisted into the Permian from the Carboniferous (Palaeodictyoptera, Protodonata, Pro- 
torthoptera, Megasecoptera), the Lower Permian fauna also included two interesting 
orders which so far as we know did not survive beyond the Paleozoic. One was the 
Protelytroptera (Fig. 102, at right), small, beetlelike creatures, with the fore wings 
modified to form covers for the thorax and abdomen. The other order was the Pro- 
toperlaria, hexapods which closely resemble the existing stone flies and were probably 
ancestral to them (Fig. 104). The nymphs (Fig. 105, at right) were aquatic and 
breathed by means of nine pairs of lateral abdominal gills, the vestiges of which were 
present in the adults. Almost all of the recent stone flies lack abdominal gills, but 
in the more generalized species (family Eustheniidae) the nymphs have them on 
the first five or six abdominal segments and vestiges are visible in the adults. The adult 
Protoperlaria were so similar to some of the Protorthoptera that it may be said truly 
that the former are Protorthoptera which have become aquatic in the nymphal stages. 

Three other recent orders make their appearance in the later Permian, one of 
these being the true beetles (Coleoptera), the second the thrips (Thysanoptera), and 
the third the true stone flies (Perlaria) . 

There are two important problems toward whose solution knowledge of the 
Paleozoic insects has contributed. One of these is the origin of the wings, for, unlike 
those of birds and other flying animals, they are not modified fore limbs but entirely 
new structures. The second and third segments of the thorax of living insects bear 
the wings, the first segment or prothorax being reduced or modified for some special 
purpose. But in the Carboniferous insects, generally speaking, and in some of the 
Permian ones (Protoperlaria, Protorthoptera), the three thoracic segments are nearly 


alike in form and structure, and, what is more significant, the first supports a pair 
of lateral outgrowths resembling small wings. The constant appearance of these 
wing-flaps in so many unrelated Paleozoic groups indicates that such structures were 
probably present in the progenitors of the Pterygota. It seems likely that the functional 

FIG. 104. Reconstruction of Lcmmatophora y a protoperlarian from the 
Lower Permian of Kansas, with membranous prothoracic lobes, and vestiges 
of gills along the sides of the abdomen. The venation is like that of the 
Recent stone flies. Sc, subcosta; Ri, radius; Rs, radial sector; MA, anterior 
media; MP, posterior media; CuA, anterior cubitus; CuP, posterior cubitus; 
lA, 2A, first and second anals. From original drawing by F. M. Carpenter. 

wings arose as such lobes and that at one stage in the history of the flying insects 
each segment of the thorax possessed similar lateral outgrowths. For some reason 
the pair next the head did not develop into useful wings and have disappeared entirely 
from Mesozoic and later forms. 

The other important question upon which Paleozoic history sheds some light is 

FIG. 105. At left, Clatrotitan, a protohemipteron from the Triassic of New 
South Wales, possessing an extraordinarily large sound-producing organ on 
the fore wing. The wing shown in the photograph is about six inches long. 
From C. Anderson. At right, restoration of a Protoperlarian nymph, from the 
Lower Permian of Kansas. These nymphs were aquatic and possessed lateral 
abdominal gills. Original drawing by F. M. Carpenter. 

FIG. 106. An unusually well-preserved dragonfly of the genus Protolindenia 
from the Upper Jurassic at Solenhofen, Bavaria. From a photograph 
specimen in the Carnegie Museum, Pittsburgh, Pa. 


FIG. io6A. A neuropteran, Lithosmylus columbianus (Cockerell), distantly 
related to the modern lace-wings. The specimen is from the Miocene of Colo- 
rado, and now in the Museum of Comparative Zoology. Much enlarged. 
Courtesy of Frank M. Carpenter. 


the origin of a complicated life history or metamorphosis. Some hexapods for 
example, the grasshoppers have much the same general structure from the time 
they issue from the egg until they reach maturity; their wings appear at an early 
stage in their ontogeny and gradually increase in size. Such a development is termed 
an incomplete metamorphosis (hemimetabolism). Others, such as the moths, differ 
in the early stages from the adult; their wings develop inside the body and appear 
only during a quiescent stage, known as the pupa. This is complete metamorphosis, 
or holometabolism. So far as is known, all the insects of Carboniferous developed by 
incomplete metamorphosis. In the Lower Permian there are two orders (Mecoptera 
and Neuroptera) which are now holometabolous, and the presence of true larvae 
in those strata shows that complete metamorphosis had been acquired. Since that 
time such forms have gradually become more abundant at the expense of those 
with incomplete metamorphosis, until 88 per cent of the existing species have holo- 
metabolous development. Although no Carboniferous insects are known to possess 
complete metamorphosis, the fact that in the Lower Permian two orders are in that 
category suggests that this form of development originated during the earlier period 
and may have been due to late Paleozoic reduction in temperature. 

Comparatively little is known about Mesozoic insects, especially those of the 
Triassic and Cretaceous. This is unfortunate because it is difficult to connect the 
modern fauna with that of the Permian. Enough early Mesozoic species have been 
found, howeverj to show that the fauna of that time was decidedly more advanced 
that that of the late Paleozoic. One insect, recently found in the Mid-Triassic of 
Australia, had a sound-producing organ on its wing much larger than that in any 
living form (Fig. 105). The sound produced by such individuals could have been 
heard for a great distance. It is interesting to reflect that they probably made more 
noise than any other animals of the time, for neither birds nor mammals are known 
to have existed then. Whether contemporary amphibians croaked or reptiles grunted 
is unknown. 

Five modern orders appeared for the first time during the Mesozoic; one, the 
Heteroptera (true bugs), is found in the Triassic; the other four, Trichoptera (caddis 
flies), Dermaptera (earwigs), Diptera (true flies), and Hymenoptera, in the Jurassic. 
Since the oldest known Hymenoptera include some parasitic forms, it is obvious 
that the group is much older than the Jurassic. Curiously, the more highly organized 
members of this group did not appear until the Tertiary. Beetles were probably the 
most abundant insects of the Mesozoic; Orthoptera were common; and dragonflies 
(Fig. 106) were undoubtedly much more numerous than they are at present. The 
recent discovery of insect-bearing Cretaceous amber in northern Canada bids fair 
to extend our knowledge of late Mesozoic forms greatly. 

Tertiary insects are abundant, especially those of the Oligocene and Miocene. 
The Lepidoptera (butterflies and moths) first appear in the Eocene, although the 


order is probably much older than that. The oldest termites are in the Baltic amber 
(Oligocene), and this is also true of certain obnoxious forms, such as the fleas. The 
modernization of the insectan fauna in the Tertiary was completed by the arrival on 
the stage at this time of the so-called social insects, bees, wasps, and ants. The ex- 
traordinarily rich faunas of the Baltic amber and of the Miocene shales of Florissant 
have enabled W. M. Wheeler, C. T. Brues, F. M. Carpenter, T. D. A. Cockerell, and 
others to describe the habits, as well as the morphology, of many of these ancient 
Hymenoptera. The results are summarized in a masterly way in Professor Wheeler's 
Social Insects. 

Ants, among the most perfect social forms, are fortunately about the most abun- 
dant Tertiary fossils of this class. They are not common in the Eocene, but Wheeler 
was able to study 11,711 specimens from the Oligocene, and Carpenter almost as 
many from the Miocene. According to Wheeler the "amber" ants are as highly 
specialized as those of the present day. Six of the eight modern subfamilies are repre- 
sented in the fauna. That social life was then highly organized is shown by the fact 
that the castes were as fully differentiated as at present, for in some genera there 
were workers of various forms. Even in those days the ants had their "cows," aphids 
or plant lice; blocks of amber have been found enclosing both the domestic animals 
and their warders. There is also some evidence that the ants welcomed certain beetles 
as guests in their nests. Only 44 per cent of the genera of amber ants seem to be ex- 
tinct; eight of the species are practically indistinguishable from forms now living. 
So far as can now be seen, this group had reached its objective as early as the Oligo- 
cene and has made no progress since. Various sorts which lived in the north temperate 
zone in Mid-Tertiary times later retired to the tropics, the present stronghold of the 
Formicidae. Inferences might be drawn from these two facts. Will Homo sapiens 
eventually reach a culmination of physical evolution, cease to struggle with wintry 
weather, and lie down beneath a palm, each queen to be served by thousands of sterile 
workers? If so, will the race maintain itself as the ants have done, or shall we become 
as rare as tapirs? 


Be commonplace and creeping, and you attain all things. 

Beaumarchais, Barbier de Seville 

A history of Mesozoic times related by some mighty dinosaur to an admiring 
group of smaller reptiles would probably have been in the main a recitation of how, 
from an obscure nativity, their race had made themselves masters of land, sea, and 
air. He might have mentioned the amphibians to substantiate the claim of progress, 
recalling the lowly stock from which the overlords had sprung. Even the birds may 
have won recognition as close, though unimportant relatives, but it is doubtful if 
the obscure little furry animals which lurked about in the underbrush would have 
been mentioned at all. Probably they had never been noticed. Although the history 
of the mammals began almost as early as that of the dinosaurs, it was not a startling 
or colorful story so long as the reptiles retained dominion. Apparently the Mesozoic 
world was exactly fitted for reptilian life. Warmth of blood and the protection of 
hair were no particular advantage in times when nearly featureless continents per- 
mitted the extension of subtropical conditions much further toward the poles than 
at the present day. When food was plenty, neither activity nor intelligence aided 
animals greatly, for there was little competition. Life was relatively easy; only with 
struggle is there progress. 

Creeping from their cradle in southern Africa during the late Triassic, the mam- 
mals spread first into Europe and later, in the Jurassic, reached Mongolia and America. 
The Mesozoic fur-bearers, although small and unimportant in their day, are of great 
interest from the evolutionary standpoint. 

Mammals are warm-blooded creatures which suckle their young. Most of them 
have a covering of hair or, in the case of the naked ones, such as elephants and whales, 
some trace of hair, at least on the young. Although some have a scaly covering, a 
hair is not a modified scale but an entirely different structure, confined to this group. 
The presence of a muscular diaphragm separating the thoracic and abdominal parts 
of the body cavity, a larynx, an external ear, a four-chambered heart, and other 
features peculiar to the class, are important specializations but not of much assistance 
to the paleontologist. The skeleton, however, shows many notable characteristics 
which allow ready identification. Only the more conspicuous need be enumerated 

The skull differs from that of other vertebrates in being more compact and having 


fewer bones. Each ramus of the lower jaw consists of a single bone, in contrast to 
the numerous elements of the reptilian jaw. It articulates with a process of the squa- 
mosal, for the quadrate, the articulatory bone of the lower vertebrates, has been drawn 
into the ear, where it forms the incus. There is a single temporal arcade, but instead 
of being composed of the quadratojugal and jugal, as in the reptiles, the posterior 
part is a projection from the squamosal, whereas the anterior portion may be the 
jugal or a process of the maxillary. At the base of the skull are two occipital condyles 
laterally placed, as in the Amphibia. Finally, the teeth of most mammals are hetero- 
dont, subdivisible into groups known as incisors, canines, premolars, and molars. 
The incisors are in the premaxillae, the others in the maxillae of the cranium. The 
incisors and, in most, the canines are single-rooted, the premolars double-rooted, the 
molars generally double or, in the upper jaw, triple-rooted. With the single exception 
of some of the ceratopsian dinosaurs, all vertebrates other than the mammals have 
single-rooted teeth. 

Nearly all mammals have seven cervical vertebrae, the principal exceptions being 
certain sloths with six, eight, or nine. The first of the series, a large, ringlike bone 
which receives the occipital condyles of the skull, is known as the atlas; behind it is 
the axis, with a peglike anterior process on which the atlas rotates. The trunk verte- 
brae are readily separated into two regions, an anterior series of dorsals, with ribs, 
and the posterior ribless lumbars. The faces of the centra of the individual vertebrae 
in these regions are flat, those of the neck opisthocoelous. The limbs are straighter 
and carried more beneath the body than is commonly the case with Amphibia or 
Reptilia. The coracoid element of the pectoral girdle is present as a distinct bone only 
in the monotremes, being vestigial and fused to the scapula in others. All of the 
elements of the pelvic girdles of most are so fused together in the so-called "innomi- 
nate" bone that it is not easy to make out the exact boundaries of ilium, ischium, and 
pubis. And, lastly, the typical phalangeal formula is 2, 3, 3, 3, 3, although one or more 
fingers and toes have been lost by many mammals. 

Present-day mammals are generally subdivided into three groups. The first is a 
small one, the monotremes, represented by the duckbill and the spiny anteater of 
Australia. These animals betray their primitive position by their reptilian habit of 
laying eggs, and the equally reptilian position of the limbs and structure of the 
shoulder and pelvic girdles. The pectoral girdle, particularly, retains the numerous 
bones of the lower animals. Adult monotremes are toothless, but the young have 
incompletely formed vestigial teeth. 

The young of the marsupials, a second and more important group, are feeble at 
birth and are carried about for some time in the pouch of the mother. An extra pair 
of bones on the pubes, known as epipubic or marsupial bones, present in no other 
mammals except the monotremes, differentiates the pelvis from that of the placentals. 
Other unusual features are shown in the teeth, which are peculiar in that all execpt one 


in each jaw are permanent; the only replaced teeth are the last premolars. Curiously, 
marsupials have three premolars and four molars, whereas the higher mammals have 
four premolars and three molars in each jaw. The modern marsupials are chiefly 
Australian, the exceptions being the opossums of North and particularly of South 

The third and largest group of mammals includes those which while within 
the body of the mother are nourished through a spongy mass known as the placenta, 
whence they are called the Placentalia. With rare exceptions, these have forty-four or 
fewer teeth. If the typical number is present, there are in each jaw three incisors, one 
canine, four premolars, and three molars. For convenience, the dentition is written as 
a formula in which the teeth of one upper and one lower jaw are included. Thus, for a 
primitive mammal, one would write: i-f-j c 4-> P4> m 4> but for ourselves: i^> c -f> 

3143 21 

p -|-, m -jp since we have on each side only two incisors, one canine, two premolars, and 
three molars. The formula must be multiplied by two to show the total number of 
teeth. Another feature of the dentition of the placentals is that there is a milk series 
consisting of the incisors, canines, and premolars, all of which are replaced by perma- 
nent teeth. 

Because of the presence of two occipital condyles, it was at one time supposed 
that the mammals arose from the Amphibia, but with a fuller knowledge of the 
theraspid reptiles of the Triassic of South Africa it has become obvious that the latter 
is the group from which the class was derived. 

The Theraspida, descendants of the pelycosaurs, include several groups, the most 
important of which is that of the carnivorous theriodonts, probably ancestral to the 
mammals. They were comparatively small animals which walked and ran rather 
than crawled; most of them had sharp conical teeth. One of the most obvious mam- 
mal-like characteristics is the division of the dental series into incisors, canines, and 
molars. This is especially well shown by Cynognathus or Thrinaxodon (Fig. 107), 
which has tricuspate posterior teeth. As in mammals the canines are the anterior 
teeth of the maxillaries, the incisors are in the premaxillaries, and the cheek teeth are 
the largest. The formula is variable, but some have i-1 c-j-> P + m-f-X 2 = 54. One 
of them, Sesamodon, said to be more like the mamals than any other member of 
the group, has the formula i-i- 5 c-L, p -f m^., or only two more than the placental 
mammal. Other mamalian features of the dentition 'are the bite of the mandibles 
inside the upper teeth, with the lower canine crossing in front of the upper one, and 
the method of replacement of teeth. Broom has shown that in some species at least 
the later teeth come up in the old sockets directly beneath the milk teeth instead of 
between the teeth, as in other reptiles, and there is some evidence that the posterior 
cheek teeth are not replaced. Other mammalian characteristics shown by one or 
another of the theriodonts are: the presence of a partially divided occipital condyle, 


or of three condyles, the loss of the median of which would produce a mammalian 
condition; a mammal-like palate which pushes the posterior nares far back; a zygo- 
matic arch formed by squamosal and jugal; a lower jaw formed mainly by the dentary; 
seven cervical vertebrae; fused pelvic bones; a mammal-like carpus and tarsus; and 
the phalangeal formula 2, 3, 3, 3, 3. 

Although so similar to mammals, the theriodonts still retained many reptilian 
characteristics. The brain was small, with a pineal opening present in most; the 
lower jaw did not articulate with the squamosal but with a small quadrate; and a 
vestige of the quadratojugal still remained. In the most primitive ones a large post- 
orbital bone extended backward from the eye along the margin of the parietal, thus 
bordering the temporal opening as in the typical Synapsida. Prefrontals were also 
present. Although the dentary made up the greater part of the lower jaw, other 
elements were present on the inside and at the posterior end. 

FIG. 107. Thrinaxodon, a small Triassic mammal-like reptile: pa, parietal; 
po, postorbital; pf, prefrontal; n, nasal; /, lacrimal; mx t maxilla; /, jugal; 
sq, squamosal. About natural size. Redrawn after W. K. Gregory. 

Recently Broom has described the remains of an important reptile from the 
Upper Triassic of South Africa. The specimens are incomplete, but they appear 
to furnish a needed connecting link. The best-known genus is Ictidosaurus, a small 
creature about the size of a rat. The skull is almost entirely mammalian, the pre- 
frontals, postorbitals, and pineal foramen being absent, and two occipital condyles 
present. The single reptilian feature is the presence of a small quadrate bone with 
which the lower jaw articulates. The lower jaw likewise retains one reptilian bone at 
the posterior end where it articulates with the cranium. Except for this articulation 
of the jaw, Ictidosaurus is, so far as is now known, a mammal. It is not at all likely, 
however, that it was a direct ancestor of any later mammal, for its small conical teeth 
are less mammal-like than those of the cynodonts. It is hoped that the unearthing of 
this new type will be followed by the recovery of the bones of many of its relatives, 
for among them we expect to find creatures more nearly in the direct line of ancestry. 

The mammals found in the Mesozoic strata are known chiefly from minute 
separate teeth and from jaws less than an inch long. Such fossils are rare in the Triassic 
but fairly common in certain places in the Upper Jurassic. Many species have been 
described, although not much can be learned about any one of them. No whole 


skeleton has yet been found in Mesozoic rocks; recently a few skulls have been 
collected from the Cretaceous of Mongolia. 

The oldest remains of mammals are found in strata of Upper Triassic age in 
South Africa and Germany, but the specimens are so few in number that knowledge 
of these earliest representatives of the group is meager. A single incomplete skull 
of Tritylodon longaeus (Fig. 108, at left) was long ago described by Sir Richard 
Owen from the Karroo formation (Triassic) of Basutoland. At first supposed to be 
a mammal, it was later considered one of the mammal-like reptiles; now opinion 
seems to have returned to a belief in the original determination. From the Keuper 
of Germany small isolated teeth have been obtained. All these specimens represent 

FIG. 108. At left, the Triassic allothere, Trytylodon. i 1 , i 2 , i 8 , incisors; 
M 1 to M 7 , molars; Pmx, premaxilla; Pal, palatines. One-half natural size. 
From G. G. Simpson, Mesozoic Mammals, by permission of the Trustees of 
the British Museum (Natural History). At right, palatal and lateral views of 
the skull of Taeniolabis, a Paleocene allothere from New Mexico. One-sixth 
natural size. From Granger and Simpson. 

an ancient group which have peculiar cheek teeth with many tubercles on the crowns, 
a characteristic which suggested the name Multituberculata, by which they are com- 
monly known, but the present scientific usage is to call them the Allotheria. Al- 
though they represent a line of mammalian evolution .which cannot at the present 
time be connected with that of the higher representatives of the group, they are en- 
titled to some consideration, since they spread through Europe, central Asia, and 
North America. They became increasingly abundant in Jurassic and Cretaceous 
times and reached their maximum in size in the oldest period of the Tertiary, the 
Paleocene, just before their complete extinction. No entire skeleton of a member 
of this group has yet been discovered, most species having been founded on teeth. 
The greater part of a skull has been described from the Upper Cretaceous of Mongolia, 
and good ones, with fragments of the skeleton, from beds of Paleocene age in central 
Montana and New Mexico (Fig. 108). 


The most characteristic feature of this group is the presence of large, elongated 
cheek teeth, with numerous tubercles on the crowns. The lower ones have two, the 
upper, three longitudinal rows of low blunt cones, obviously better adapted for 
crushing than for cutting or tearing food. Canines are lacking, but the incisors are 
elongate, erect in some forms, in others procumbent, projecting far forward. Com- 
paring the dentition of the best-known species with that of modern animals, it is 
inferred that many of them were rodentlike in their habits, probably using the erect 
incisors for gnawing or pulling the bark from cycads or other plants. Those with 
procumbent incisors may have fed on fruits, nuts, eggs, insects, and such small ani- 
mals as they were able to catch. Like some modern rodents, they were more or less 
omnivorous. All were small, the largest the size of a modern beaver, but most of 
them comparable to mice. 

It is obvious from the lack of canines and the presence of large, peculiar molars 
that all were specialized, in spite of their great antiquity. This impression is borne 
out by what little is known of the structure of the skull and other bones. The rela- 
tionships, so far as they can be determined, appear to be with the marsupials, although 
there is some indication that they may have been allied to the early monotremes. In 
any event, they were not progenitors of the modern mammals but form a subclass 
of their own. 

The Jurassic strata have furnished three sorts of jaws and teeth of a primitive 
nature. The more common are the Triconodonta (Fig. 109, at right), a name sug- 
gested by the fact that each cheek tooth has three fangs in a line, the median high, the 
anterior and posterior ones subordinate to it. Triconodonts have been found in 
England and North America, but the number of species is not great. They were 
small, conjectured to have been about the size of mice. Little is known of the skele- 
ton, since the best material so far found retains no more than lower or upper jaws 
and parts of the skull associated with the palate. Most of them had numerous teeth, 
the lower jaw commonly carrying a canine, four premolars, and from three to six 
molars. Not many specimens retain the incisors, but such as have them exhibit three 
or four small, slender, conical teeth on each side of the jaw. All the cheek teeth, and 
in most the canines, are double-rooted, although the roots of the premolars are not 
in all cases fully divided. 

The general appearance of this dentition is very much like that of some of the 
cynodont reptiles, especially such a one as Cynognathus. The triconodonts were, 
however, true mammals, as is shown by the double-rooted cheek teeth and the fact 
that the dentary makes up the whole of the lower jaw. Intermediate structurally 
between the cynodonts and the triconodonts are the mammal-like reptiles from the 
Upper Triassic of North Carolina. Two little jaws, each less than an inch in length, 
were found long ago by Ebenezer Emmons and described by him as Dromatherium 
sylvestre (Fig. 109, at left). One of them was later removed to another species and 


genus. Both resemble triconodonts, and until recently they were supposed to be the 
most ancient American mammalian remains. W. K. Gregory, and lately G. G. Simp- 
son, have shown, however, that there is reason to doubt this identification, for the 
jaws display some reptilian and no positively mammalian characteristics. The teeth 
are similar to those of the triconodonts, Dromatherium having three incisors, a canine, 
and ten cheek teeth, the last seven of which are definitely three-cusped. Detailed 
study, however, shows that the cheek teeth are not actually double-rooted, the root 
being merely laterally compressed, with a median depression which does not fully 
divide it. They are therefore intermediate in structure between the single-rooted 
ones of the cynodonts and the double-rooted ones of the triconodonts. Moreover, 
although the jaw appears to consist of only one element, the dentary, there is reason 
to believe that a part is missing from the posterior end and that another small bone 

FIG. 109. At left, Dromatherium , a mammal-like reptile from North Caro- 
lina. One and two-thirds natural size. From G. G. Simpson. At right, a 
reconstruction of Trioracodon, a Jurassic triconodont One and six-tenths 
natural size. From G. G. Simpson, Mesozoic Mammals, by permission of the 
Trustees of the British Museum (Natural History). 

was present at the articulatory angle. The triconodonts seem to have been derived 
directly from cynodonts similar to Thrinaxodon, but they perished at the end of the 
Jurassic without giving rise to any other group. 

Modern marsupials are of two kinds. Some, the kangaroos and their allies, 
have two prominent incisors, whence they are called the Diprotodonta; others, 
more primitive, with numerous small incisors, are the Polyprotodonta. A well- 
known example of the latter group is the American opossum. Opossums like 
modern ones existed in late Cretaceous times, and animals that seem to be closely 
related are found in earlier Cretaceous and Jurassic strata. They are characterized 
by the possession of numerous teeth, more or less like those of later marsupials 
and placentals. The cheek teeth of members of this group are approximately tri- 
angular in form, each with three principal cones, one at each corner of the triangle. 
The same pattern is found in the early opossums and insectivores, the latter the 
most ancient of the placental mammals. Fossils recently found in the Cretaceous 
of Mongolia have been definitely proved to be the remains of placentals, and, 


in fact, of true Insectivora. Because of the likeness of the dentition, the insecti- 
vores and the polyprodonts and, inferentially, all the marsupials and placentals are 
now supposed by paleontologists to have had a common ancestry in the oldest known 
mammals possessing triangular teeth. These ancestral animals, the Trituberculata 
or Pantotheria, made their first appearance in England in Mid-Jurassic times. Their 
oldest representative, Amphitherium, had numerous teeth, one species having a for- 
mula which, by reduction, could be modified to produce either the marsupial or the 
placental dentition. This 'scheme of derivation of the marsupials and placentals from 
a common ancestor, rather than one from the other, gets over one or two manifest 
difficulties inherent in the older idea that the placentals, more specialized by virtue 
of their longer period of gestation, must have been derived from the marsupials. 
Although obviously in many respects more primitive than the placentals, marsupials 
are more specialized in that the milk dentition is so reduced that only one tooth on 
either side is shed, whereas in the placentals all but the true molars are replaced. 
This theory explains why marsupials, although not descended from the placentals, 
show in one modern group a sort of rudimentary placenta. 

The trituberculates are, therefore, highly important, for they are probably an- 
cestors of most if not all living mammals. Although their study has not yet given a 
clue to their own ancestry, they seem to have been derived from some sort of therio- 
dont reptile. During later Jurassic times this group spread to North America and 
East Africa, where their remains occur in strata associated with those which have 
furnished the bones of the great sauropod dinosaurs. All were small long-jawed 
creatures outwardly like mice or rats. Their incisors were conical, their canines sharp, 
double-rooted cones, and the premolars small, sharp, piercing or bladelike cutting 
teeth. The molars were characteristic, mostly of triangular form, with four or more 
cusps on the crowns of the upper ones, three on the lower. Primitively there seem to 
have been four premolars and eight molars, but some show considerable reduction 
from this large number of cheek teeth. It is unfortunate that so little is known of the 
skeleton of these animals. So far as can be judged from the jaws, they were par- 
ticularly well equipped for catching and devouring insects and other soft-bodied 
invertebrates, or any vertebrates smaller than themselves. They probably lived chiefly 
upon flesh, although they may have eked out their diet by eating such fruits or seeds 
as were available. They vanished with the Jurassic, but the presence of their triangular 
molars in the marsupials and insectivores of the Cretaceous is evidence that they were 
transformed into higher types of mammals. 

The triconodonts and the trituberculates are the most abundant of the three groups 
of mammals which made their first (and last) appearance in the Jurassic. The third 
group, known as the symmetrodonts (Fig. no), is characterized by a simple triangu- 
lar molar that was at one time considered to be intermediate between those of the 
two sorts already described. These teeth are primarily simple cones, but each bears 


two secondary cusps, which appear on the outer side of the upper teeth and on the 
inside of the lower ones. The teeth are much like those of the triconodonts, except 
that their form is triangular. It was natural that paleontologists a couple of genera- 
tions ago should believe that the triangular tooth of the trituberculate arose by a 
rotation of the subordinate cusps of the triconodont, outward in the upper jaw and 
inwsgrd in the lower. But since no intermediate forms have been found, this idea 
has been discarded by those who have recently studied the evolution of teeth. Appar- 
ently the symmetrodonts were never numerous, and so little is known of them 
that it is impossible to say whether or not they v^ere closely related to other Jurassic 

** i# 

FIG. no. Diagrams to show, in lateral and crown views, the possible 
occlusal relationship of symmetrodont teeth. From G. G. Simpson, Mesozoic 
Mammals, by permission of the Trustees of the British Museum (Natural 

FIG. in. The jaw of the Cretaceous opossum, Eodelphis. After W. D. 

The oldest true marsupial so far found has been named Eodelphis (Fig. in). 
It was described from jaws collected in the Belly River formation of the Upper Cre- 
taceous. Although an opossum, it cannot have been a direct ancestor of the modern 
ones, for it has one less tooth in each jaw. It is, however, sufficiently like the modern 
representatives of its group to prove that, so far as lineage is concerned, the opossums 
carry the true blue blood, no other modern mammalian family being traceable so 
far back as the Cretaceous. 

The expeditions conducted by Dr. Roy Chapman Andrews in central Asia 
have evoked tremendous public interest and achieved lasting fame by the discovery 
of dinosaur eggs. But nothing they have produced can vie in scientific importance 
with a small part of the collection made in 1925, a few fossils in which the public 


has evinced no interest whatsoever. These are seven small, incomplete skulls, the 
oldest known remains of placental mammals. They have been described by Dr. 
George G. Simpson, who has contributed greatly to our knowledge of Mesozoic 
mammals by redescribing all the known species of the world. He had the courage 
to undertake the study of a group that had previously defied analysis, and developed 
a technique which enabled him to bring order out of chaos. Four genera and five 

FIG. 112. Skulls and restorations of the heads of two Mongolian placenta Is. 
The one at the left may have been ancestral to some sorts of modern insecti- 
vores. The other has a more carnivorous type of dentition, suggesting rela- 
tionship to the later creodonts. One-half natural size. From W. K. Gregory, 
in the Scientific Monthly, by permission of the Science Press. 




FIG. 113. Diagram to show the probable relationships of the various groups 
of Mesozoic mammals. From G. G. Simpson. 

species are represented by these seven specimens. All appear to be related to the 
modern Insectivora, a group generally accepted as the most primitive in its subclass. 
Two families, showing divergent evolutionary trends, are described. The teeth of 
one are in a continuous series, the canines enlarged and single-rooted, the premolars 
trenchant, and the triangular molars sharp. Although the members of this family 
are insectivores, their teeth suggest carnivorous habits, and it is believed by Simpson 
and Gregory that the oldest Tertiary carnivores, the creodonts, as well as many of the 
later Insectivora, may have sprung from this stock. The other family represents a 
somewhat specialized group of insectivores, with the lateral upper and median 


lower incisors elongated. There is a somewhat long toothless portion of the jaws 
behind the incisors, followed by triangular cheek teeth. Although not closely re- 
lated, the animals of this group may have been allied to the ancestral line of the 

These few skulls, obtained by a party working under difficult conditions, hin- 
dered by lack of time and by inhospitable surroundings, are only a sample of what 
the Mongolian Cretaceous may have in store for paleontologists. They encourage 
us to expect wonderful additions to our knowledge as exploration proceeds. Fossils 
of this region may eventually make it possible to write a satisfactory introductory 
chapter to the history of mammals. 

At the present time six groups of Mesozoic mammals are known. Most abun- 
dant are the multituberculates, which certainly arose from some theriodont reptile of 
the Triassic. They survived to the Eocene but did not give rise to any other group. 
Next are the triconodonts, probable offspring of the cynodonts, possibly of a form 
not unlike Cynognathus. They were little mouselike creatures, which, like the simi- 
lar symmetrodonts, did not outlive the Jurassic and left no descendants. More im- 
portant are the trituberculates, Pantotheria, known from the Jurassic only, but an- 
cestors of the modern mammals. From them came the primitive marsupials, the 
opossums of the Cretaceous of America, and the primitive placentals, the insectivores 
of the Cretaceous of Mongolia. Asia, long traditionally held to be the birthplace of 
man, may have been that of the placental mammals. 


What! will the line stretch out to the crack of doom? 

Macbeth, Act IV, scene i 

The later part of the Mesozoic witnessed a transfiguration of the earth. Vast 
shallow seas disappeared from the interiors of the continents and receded from their 
borders; great mountain chains surged up into billowlike crests, and the courses of 
ocean currents were altered. This geographic revolution changed climatic conditions 
in various regions and brought about a new vegetation. As early as the Lower Cre- 
taceous new trees appeared, trees which bore conspicuous flowers, shed their leaves 
annually, and yielded fleshy fruits of great food value". By the end of that period 
deciduous angiosperms had replaced, in many areas, the pines and other evergreens, 
the ginkgos and cycads, which had been the common trees of the Mesozoic. During 
the early Tertiary, as the climates of the continental interiors became drier, incapable 
of sustaining a population of larger plants, grasses appeared. These changes had a 
profound effect upon all the animals of the globe. Exactly how and why are still 
mysteries, but life on both land and sea showed considerable alteration. The great 
reptiles of the land and the humble mollusca of the sea found the times equally diffi- 
cult. Some survived this period of stress and change; others vanished completely. 

It was a new earth that the little mammals inherited in the Tertiary, an earth 
furnished with vegetable food in greater abundance and variety than had ever 
previously existed. Danger from reptiles other than crocodiles, snakes, and turtles 
had passed; the overladen table called for eaters of fruits, nuts, leaves, and grasses. 
The world was, as never before, fitted for vegetarians. It is not surprising, then, 
that the mammals began to thrive, to wax strong, and to differ one from another. 
As rTas been seen, the first placentals, the insectivores of the Mongolian Cretaceous, 
had tirangular cheek teeth with high cones on the crowns. When the jaws were 
closed,, the upper and lower teeth sheared past one another. Such jaws would be 
useful in catching and cutting insects, worms, small mammals, fruits, seeds, and suc- 
culent plants. One infers, therefore, that the early insectivores were really omniv- 
orous. They are called insectivores because they are most closely related to such 
present-day mammals as live chiefly on insects, not because of their own diet. 

Although only a few Cretaceous placental mammals are now known, there must 
have been hundreds of species and myriads of individuals of the little creatures. 
Modern insectivores, which retain to a surprising degree the primitive dentition and 


skeletal characteristics of their distant ancestors, occupy many habitats. Many, such 
as the hedgehogs, are terrestrial; some, like the moles, burrow; but others, the shrews, 
dwell in trees, and still others are semiaquatic, subsisting largely on fish. It is prob- 
able that the Cretaceous ancestors were equally versatile, seeking food in various 
situations. As time went on, the different clans became accustomed to particular 
types of food. Some sought it constantly in trees, their habits of climbing and jumping 
gradually bringing about changes in the skeleton, lengthening the arms and de- 
veloping the hands and feet for grasping. The more ambitious of them learned to 
glide, and finally to fly. Others ran about on the ground eating leaves and twigs. 
In these, constant running developed the central toes, enabling them to forsake the 
flat-footed gait, lengthened the legs, and eventually produced hoofs. Still others fed 
on bark; gnawing strengthened the front teeth, whereas the molars became grinders. 
These mammals sought for roots and their feet gradually changed into implements 
for digging. But even with an abundance of vegetable nourishment the old Mesozoic 
blood-lust lingered. Flesh is, after all, the most concentrated of foods. Those who 
eat it have time for something more than feeding time -to sleep and clean their 
coats, to make themselves attractive and to purr, time to gain strength to fight and 
make themselves masters of the world. The sharp teeth of the early placentals indi- 
cate that all would eat flesh when they could get it. There was not enough for all, 
so the bolder, the larger, the cleverer, got what there was, and finally became solely 
beasts of prey. 

These changes were not instantaneous but gradual. The fossils are sufficiently 
abundant to allow some lines to be traced; in other groups the records are so scanty 
that with our present limited knowledge it seems almost as though they had been 
specially created. Our experience of the discovery of one connecting link after an- 
other leads us to believe, however, that sooner or later the whole lineage will be 

The oldest Tertiary mammals are found in Paleocene formations, which have 
been productive of fossils at only a comparatively few localities. The Paleocene has 
been to students of the Tertiary what the Cambrian is to geologists in general. It 
was the last epoch to be recognized, and only recently has its distinctive fauna come 
to be known. Within the past decade or two many new fossils have been described, 
but, as is the case with the Cambrian, most of them from incomplete, unsatisfactory 
material. The three classic localities have been the San Juan basin of northwestern 
New Mexico, Cernay, near Rheims, northeast of Paris, and the Fort Union strata 
in central Montana. The last of these has recently yielded numerous fossils, and the 
productive area has been traced through southern Montana into northern Wyoming. 
Paleocene fossils are now known from various outcrops from Alberta to New Mexico, 
and one hundred and sixteen genera of mammals have been described from them. 
One of the most important of the newer discoveries was that of eleven genera in 


strata of this age in Mongolia. Interesting, but less significant, is the presence of twenty- 
one genera in central Patagonia. 

These data are from a census published by Dr. G. G. Simpson in 1936, and are 
a revelation to one who has not kept abreast of recent developments. They show that 
sixteen of the twenty-eight orders of Tertiary and recent mammals were then in 
existence. Three, the Allotheria (Multituberculata), Marsupialia, and Insectivora, 
survived from the Cretaceous. The remainder are new arrivals on the scene. Except 
for the first two orders listed above, all are placentals, a fact that suggests an extremely 
rapid expansion in that group. Where did the differentiation take place? The multi- 
tuberculates and marsupials had been in North America since the Upper Jurassic and 
the Upper Cretaceous respectively, but the placentals are all new to this continent. 
Since the oldest placentals so far found are Mongolian, paleontologists appear to have 
jumped to the conclusion that northern central Asia was their cradle. This may be 
true, but does the present evidence warrant such a belief? It may be that North 
America was the real center of origin, for at the present time ninety-five genera of 
placentals are known here, as compared with nine in France, nine in Mongolia, and 
fifteen in Patagonia. Ten orders are represented in North America, six in Mongolia, 
and five in France. Combining France and Mongolia to represent Eurasia, there are 
seven. Only three are known from Patagonia. 

It is true that in the present state of exploration statistics do not have much 
significance. Those just quoted may be interpreted as meaning (a) that the North 
American strata have been searched much more thoroughly than those of Mongolia, 
(b) that the first great differentiation of the placentals may have taken place in North 
America, even though they originated in Asia, or (c) that the placentals actually did 
originate on the continent of North America early in Cretaceous times. 

That the last is possible seems to be supported by the following arguments, the 
first of which has already been cited and to some extent discounted. 

1. The Paleocene mammals were much more numerous and much more highly 
differentiated on this than on other continents. 

2. The most primitive of placentals, the insectivores, are much better repre- 
sented here than in Eurasia. Twenty-three genera are known from North America, 
two each from Mongolia and France. Furthermore, the primates, which stand next 
in the scale of organization, are represented by twelve genera here and only one in 

3. Such evolution on this continent was possible. Everyone agrees that the placen- 
tals descended from the pantotheres, which were just as common in North America 
during late Jurassic times as they were in Europe. They have not so far been found 
in Asia, although there is reason to believe that they once existed in that region. They 
disappeared from Europe and North America at the same time, the end of the Jurassic. 
It is as probable that their further evolution was in North America as in Eurasia 


more probable, in fact, for the greatest abundance of trituberculates is in western 
Europe, away from, rather than toward, a possible center of distribution in Eurasia. 
4. It seems to be generally agreed that both the placentals and the marsupials 
were derived from the pantotheres. The oldset known marsupials are found in the 
North American Upper Cretaceous. They are not as yet known from the Mongolian 
Cretaceous or Paleocene, although they were in France by Eocene times. Perhaps 
this merely balances the argument from the presence of Cretaceous insectivores in 

For the sake of the record it is necessary to insert here a list of the orders of 
mammals known to have been in existence in Paleocene times. They are the Allotheria, 
Marsupialia, Insectivora, Tillodontia, PDermoptera, Chiroptera, Primates, Taenio- 
donta, Edentata, Rodentia, Carnivora, Condylarthra, Amblypoda, Litopterna, No- 
toungulata, and Astrapotheria. Only the most important can be discussed. The last 
three are South American herbivores. Only one of these orders, the Notoungulata 
(Toxodontia), is represented outside that continent. It has one genus in the Paleo- 
cene of Mongolia and one in the Lower Eocene of North America. 

All of the Paleocene mammals are primitive; nearly all have five fingers and 
five toes; most have the typical forty-four teeth. The upper premolars are triangular, 
the molars triangular or quadrangular, with blunt or elongated piercing cones. The 
legs are short, the lower bones of nearly equal size, and the gait of all was digitigrade. 
The terminal phalanges are blunt, neither claws nor hoofs. The animals are small, 
the giants among them not much larger than sheep; this, however, is a considerable 
increase over any Mesozoic mammal. Although all are much alike, still there are 
recognizable differences. 

Insectivores (Fig. 115) other than moles are not familiar to the average American, 
but the common mole, although somewhat stouter-limbed than most of its allies, 
gives a fair idea of the group to which it belongs. Most of the insectivores are small, 
with slender, somewhat elongate heads. Many have a complete dentition, although 
some have lost one or more incisors and premolars. The molars of most are triangular 
and retain many primitive characteristics. The Old World hedgehogs, not to be 
confused with that prickly American rodent, the porcupine, are relatively slightly 
modified descendants of some of the Mongolian Cretaceous placentals. They have 
not always been restricted to Eurasia, for Paleocene and Oligocene representatives 
of the group have been found in North America. The oldest of North American 
insectivores are still so imperfectly known that opinions about their relationships 
differ, but some of them are thought to be allied to the squirrel-like tree shrews of 
the Indo-Malayan region. 

The rapidity of diversification among the mammals at the beginning of the 
Tertiary is made evident by the appearance of flying mammals the bats at that 
time. It is true that only a single imperfect upper jaw has yet been found in the Paleo- 


cene (this in Colorado), but more and better-preserved specimens, even those retaining 
wing bones, are known from the European Eocene. Although the pterosaurs are the 
most specialized of the reptiles and the birds the most specialized of all vertebrates, the 
bats do not represent the greatest known departure from the primitive structure of 
the mammals. That honor has been preempted by the baleen whales. But the bats 
occupy a position which is only slightly subordinate. Had not most of them retained 
their primitive insectivorous habits, they would probably have stood (morphologically) 
at the head of the list. Bats, and perhaps bats alone, can be accused of having failed 
to live up to their full opportunities. It is true that some of them have gone beyond 
most mammals in adopting the insectan habit of blood-sucking, but apparently 
they did not become vampires early in their history. At any rate, no bat has as 
yet succeeded in losing its teeth, so the group can hardly be put on the plane of the 
pteranodonts, the post-Cretaceous birds, or the whalebone whales. But bats and 
men are on their way. Another million years may see toothless man at the top, 
for he can easily dispose of bats and whales, should they threaten to take a technical 

As a matter of fact, a flying mammal is more to be expected than a flying reptile. 
As the late Dr. W. D. Matthew pointed out, there are many reasons for believing 
that the primitive trituberculate mammals, both placental and marsupial, were 
arboreal. They had two reasons for living in trees: the first, the great size of their 
reptilian neighbors, and the second, the abundance of food available to scansorial 
creatures. The dinosaurs were not competitors of such animals as could climb. 

If, as seems probable, the late Cretaceous insectivores were arboreal, it is not at 
all surprising that some of them should have learned to fly. That the ability to fly 
came about through the growth of skin between all the fingers except the first and 
second is strange, in view of what happened to the pterosaurs and birds. It might 
have been expected that one or the other of the schemes adopted by those animals 
would have been followed. But the various creatures which learned to fly were 
differently constructed at the time when the new habits were initiated. The ancestral 
pterosaur had already lost the little finger; the reptile which gave rise to the birds 
had lost the fourth as well. The bats started from scratch, with all five. In the sequel 
those apparently most handicapped fared best, probably because of feathers. 

The lemuroids were the first and most primitive of the primates, that division of 
mammals with hands and feet equipped with nails rather than hoofs or claws. The 
Paleocene representatives of the group leave no doubt of the close relationship be- 
tween them and the insectivores, for they are so much alike that specimens are dis- 
tinguished only with difficulty. The lemuroids resembled the modern lemurs in 
having long skulls but were without some of the latter 's specialization. Though 
primates were more or less common in North America during the Paleocene and 
Eocene, their later history can be traced only on other continents. Those found in 


the Paleocene were all tiny creatures about the size of squirrels or mice, with slender 
limbs, long fingers, and the thumb (pollex) and great toe (hallux) partially opposable 
to the other fingers and toes. 

Other early offshoots of the insectivores were the Tillodontia, large rodentlike 
creatures which did not survive the Eocene, and a few small creatures which are 
tentatively placed in the Dermoptera. The only living representatives of this order 
are the so-called flying lemurs of the East Indies, somewhat batlike animals which 
glide from branch to branch like flying squirrels. They have an expansion of the 
skin extending from the neck to hands and feet, and ending on the tail. Although 
their hands are enlarged, there is no indication that they represent an intermediate 
stage between insectivores and bats. It is curious that there is no record of this group 
between the Eocene and the present time. 

The creodonts are the first carnivores. Appearing in the Paleocene, only slightly 
modified from insectivores, they became more and more abundant during the suc- 
cessive ages of the Eocene but died out in America early in the Oligocene, in Europe 
in the Miocene. They are discussed further in the chapter on beasts of prey. They 
were the sole predators of the Paleocene and Eocene, the most peaceful ages in the 
history of mammals. 

The condylarths (Fig. 114) are the most primitive hoofed mammals, interesting 
because they differ so little from the contemporaneous creodonts. As with the latter, 
the limbs were short and massive, the gait semidigitigrade, the tail long and heavy. 
The Paleocene condylarths have tritubercular teeth much like those of the creodonts 
and even similar tusk-like canines, but the later Eocene representatives had quadrangu- 
lar grinding molars of the bunodont type. The principal differences of the early 
condylarths from the creodonts appear to have been the narrower skulls, weaker 
jaws, and somewhat blunter toes of the former. The two groups are so much alike, 
however, that unless material is fairly complete it is almost impossible to distinguish 
them. All this indicates that the herbivorous hoofed mammals (ungulates) arose 
from a group of insectivores closely allied to that which produced the carnivorous 
creodonts. The condylarths did not survive the Eocene; in fact, they appear to have 
died out before more than half of Eocene time had elapsed. It cannot be proved that 
they left descendants. 

The Amblypoda are another group of unprogressive hoofed mammals, important 
in Eocene times but extinct since the end of that period. The earliest representatives 
of the race were of relatively small size, about that of pigs. They have stout limbs 
and the clumsy, blunt-toed, elephantlike feet which suggested the name for the group, 
Curiously enough, the early amblypods show many of the characteristics of the creo- 
donts, the body and tail being long and catlike, the skull bearing similar tusklike 
canines. The head was narrow, however, and the molars had crowns of the grinding 
type. The teeth seem to be more fully modified for an herbivorous diet than those 


of any other Paleocene mammals. The ancestor is to be sought in some Cretaceous 

condylarth-like creature. 

The amblypods (Fig. 116) must have been abundant during the Eocene, for one 
museum has accumulated about two hundred more or less complete skeletons of 
them. This unusually large amount of material has permitted the tracing of two or 
three lines of evolution. The one best known culminated, just before the extinction 
of the race, in the largest Eocene animals, some of them with the bulk of an elephant. 
These great uintatheres had remarkable heads, with three pairs of knoblike horns 
and extremely long, saberlike upper canines. One pair of horns was on the nose, an- 
other in front of the eyes, and the last in front of the ears; it is not known whether 
their bony cores were covered with horn or with a callous skin, although probably 
with the latter. The long tusks were in all likelihood used chiefly to pull down 
branches of trees and to strip the leaves from them, although it has been suggested that 
they could have been used for fighting if the animal struck with the mouth open, after 
the fashion of a snake. 

The Amblypoda are sometimes called the dinosaurs of the Eocene because the 
brain was smaller in proportion to the total bulk than in any other living or extinct 
mammal. Moreover, its surface was smooth, a large portion of it occupied by the 
lobes devoted to the sense of smell. These creatures seem to have reached, for 
mammals, the climax of brute strength as contrasted with brain power. Their ex- 
tinction was probably due to the smallness of the brain and the great size of the 
body. They had too little intelligence to protect their young from the crafty carni- 
vores and too much bulk to be provided with food when hard times came upon them. 
Thus as early as the Eocene the dominance of Mesozoic brute force ended, and 
evolution along the line of increasing brain began. 

The edentates, chiefly South American animals, are characterized by loss and 
degeneration of teeth; such as remain are peglike and devoid of enamel. Even their 
oldest known ancestors, the Paleocene representatives of the group in North and 
South America, are surprisingly specialized in this respect. Palaeanodon did, it is 
true, have four fairly good cheek teeth with some enamel, but the Eocene forms have 
a much feebler dentition. Although they may possibly have been ancestral to the 
armadillos, there is no evidence that they had a bony armor. Taeniodonts, which are 
allied to the edentates, are more common than the latter in North America, but 
apparently their line led nowhere. 

At one stage or another during the Eocene most of the other major groups of 
mammals made their appearance. Late in the epoch the modern types of carnivores 
emerged from the creodont stock. Perissodactyls were the predominant types of 
hoofed mammals, represented by primitive horses, tapirs, and rhinoceroses among 
groups which persist to the present day, as well as by the extinct titanotheres and 
amynodonts. The artiodactyls were almost as numerous as the odd-toed mammals, 

no. 114. me primitive conayiartn, rnenacodus. Photograph by cour- 
tesy of the American Museum of Natural History, New York City. 

FIG. 115. Diacodon, an Eocene insectivore, figured to show the trituber- 
cular teeth. One-third natural size. From Matthew and Granger. 

FIG. 116. Skull, teeth, and right hind foot of Barylambda, a large Paleo- 
cene amblypod. Note the triangular teeth. The skull is about twelve and a 
half inches long. From Bryan Patterson. 

FIG. n6A. Barylambda faberi Patterson, the largest of the Upper Paleocene 
amblypods, restored by John C. Hansen of the Chicago Natural History Mu- 
seum. Courtesy of the Chicago Natural History Museum. 


including the first of the camels, of the deer, and of the peccaries, among animals 
still existing, as well as several extinct groups, most outstanding of which is that of 
the merycoidodonts. In the Eocene of Africa are found the oldest remains of ele- 
phants and whales, representatives of the latter reaching America in late Eocene 

By the end of the Eocene the major differentiation of mammals had taken place. 
During the remainder of the Tertiary, the Oligocene, the Miocene, and the Pliocene, 
mammals became more and more like those now existing. Most of the surviving 
groups belong to families which were fully differentiated during the Oligocene or, 
at latest, the Miocene. Nothing really new, with the exception of man, has appeared 
since the latter epoch. Several groups prominent in Oligocene and Miocene times 
have long been extinct. Others died out during the cold periods of the Pleistocene, 
and still others have been or are rapidly being exterminated by man. The heyday 
of mammals, other than man, has passed. Two-thirds of their history, from the Triassic 
to the end of the Mesozoic, was passed under the suppressive rule of the reptiles. 
They made little progress during that long period, but when in the Paleocene they 
inherited the earth they rapidly made full use of it. Finally one of their own number 
has decided that he alone shall dominate. In taking possession he is exterminating 
such of his brethren as cannot be brought into profitable bondage. The history of 
the mammals has been one of constant increase in use and size of brain; with man, 
brain has finally superseded all other factors in survival value. 


Nought treads so silent as the foot of Time; 
Hence we mistake our autumn for our prime. 

Edward Young, Love of Fame, Satire v 

Claws and conical teeth are the equipment of the beast of prey, be it amphibian, 
reptile, or mammal. The carnivorous reptiles inherited their dentition directly from 
the Amphibia, but since the Jurassic ancestors of Tertiary mammals had triangular 
cheek teeth, a conical or a bladelike mammalian tooth behind the canines is a spe- 
cialized, not a primitive, instrument. Few mammals have become so specialized 
as to have teeth like those of primitive flesh-eating amphibians or reptiles. The nearest 
approach is, perhaps, in the single-rooted, conical ones of certain porpoises. A few 
truly carnivorous mammals have simple peglike teeth in the anterior parts of 
the premolar or molar series, but such are commonly vestigial, practically without 
function. The primitive tritubercular dentition was about as well adapted for 
chewing animal as vegetable food; hence it is only the abundance of the latter 
which can account for the predominance of herbivores over carnivores during the 

The principal trend among the carnivorous mammals has been toward the 
evolution of progressively more efficient cheek teeth. It will be remembered that the 
triangular teeth of the lower jaws of the pantotheres slid into the reentrants between 
the upper ones, producing a shearing action. This shear, because of the shape of the 
teeth, was irregularly zigzag, and hence inefficient. The earliest carnivores, the 
creodonts, inherited- the triangular tooth, but throughout the Tertiary the flesh- 
eaters, evolving along various lines, came to have more and more bladelike cheek 
teeth. Thus the direction of cutting became less and less irregular, until in the more 
highly specialized groups the action of the jaws is like that of ordinary household 
shears. If such a comparison be made, however, it should be, perhaps, with an old 
pair, badly nicked toward the tips but retaining good cutting edges near the rivet 
which holds the blades together. Not all of the cheek teeth became modified for 
cutting, but only one or two pairs in each jaw. Anyone who has seen the family cat 
at work upon its plate of liver must have noted that most of the chewing is done 
within a limited region on each side of the mouth. The teeth which do most of the 
work have become enlarged, highly specialized, the most important of the whole 
series. These shearing organs are called the "carnassials" or "sectorials"; to the tax- 

FIG. 117. Sinopa grangers, a doglike Eocene creodont from Wyoming. 
Note the arched back, long tail, and heavy limbs with almost plantigrade 
feet. Photograph by courtesy of the American Museum of Natural History, 
New York City. 

FIG. 1 1 8. Pseudocynodictis, a slender Oligocene dog of a type about 
which there has been much discussion. Some consider it to be in the main 
line of descent of modern dogs; others think that it is the last member of 
an extinct group. From W. D. Matthew, 


Wolf \ C famiharis /Fox 









FIG. 119. The family trees of the wolf, domestic dog, fox, and bear, as 
understood by W. D. Matthew. Such family trees are subject to constant 
revision, as "new" fossils are found. Redrawn after Matthew, with modi- 
fications after F. B. Loomis. This Cynodictis is now called Pseudocynodictis. 


onomist they are of fundamental importance in the division of the beasts of prey 
into subordinate groups. 

The order Carnivora is commonly treated as composed of three lesser sections 
or suborders. The most specialized are the Pinnipedia, the seals and walruses, adapted 
for life in the sea; the members of the other two are, in the main, terrestrial. One, 
the Fissipedia, includes all the modern land carnivores, of which there are seven 
families: the dogs; the raccoons; the bears; the mustelids, including martens, weasels, 
and skunks; the cats; the hyenas; and the civets. The last two never reached North 
America, and the bears arrived only in the Pliocene; but the others have a more or 
less full record in the Tertiary of this continent. All of these animals have but a single 
carnassial in each side of each jaw. The fourth premolar of the upper series is blade- 
like, this tooth being opposed by the first molar of the lower jaw. All Fissipedia 
have less than forty-four teeth. The third and most primitive suborder is that of the 
creodonts, early Tertiary carnivores with, for the most part, the typical forty-four 
teeth of the placental mammalian dentition. The carnassials are variable in number 
or not fully developed as shearing teeth, for, as is the case with the Fissipedia, some 
creodonts were omnivorous rather than strictly carnivorous. The few known Paleo- 
cene members of the group were small; the Eocene and Oligocene were the periods 
of greatest differentiation and also of greatest size, some creodonts of the latter time 
being as large as modern wolves or bears one, Andre wsarchus^ much larger. Only 
one family survived till the early Miocene. 

Although diversified, all creodonts (Fig. 117) share certain characteristics. The 
canines are large, the premolars simple, laterally compressed, the molars fundamentally 
tritubercular but variously modified in the several families. The head, which is 
large, shows a typical characteristic of the carnivore in its width at the orbits, bulging 
zygomatic arches giving it unusual breadth. The brain case was small, and the 
sagittal and occipital crests consequently high, to afford room for the attachment of 
muscles. The body was long, and the tail long and heavy, a characteristic inherited 
by many modern carnivores. "Lauk! what a monstrous tail our cat has got!" The 
limbs were short and stout; the feet, except in specialized members of one family, 
pentadactyl and digitigrade, or semiplantigrade. 

As has already been pointed out, some of the Mongolian Cretaceous insectivores 
have characteristics which suggest the creodonts: in the Paleocene are found scanty 
remains of animals belonging to the most primitive genera of the group. The in- 
dividuals are small, and their sharp-cusped tritubercular molars recall those of in- 
sectivores. This family, the Oxyclaenidae, may be shown, when better known, to 
have been ancestral to all other carnivores. 

One family of the creodonts, the Miacidae, has the same arrangement of carnas- 
sials as the modern carnivores; hence Professor W. B. Scott's suggestion that it should 
be linked with the Fissipedia rather than with the Creodonta. The oldest members 


are found in Paleocene and Lower Eocene formations, the youngest in the Upper 
Eocene. They are the only creodonts with p - and m - functioning as carnassials. 
The brain case was larger than that of their relatives, and the feet had five toes, 
arranged in spreading fashion, each armed with a small, sharp, partially retractile 
claw. The dogs are the only modern carnivores as yet definitely connected with this 
group, but all Fissipedia must trace to it even though some of the connecting links 
are still unknown. 

Turning now to the modern terrestrial carnivores (Fissipedia), no attempt will 
be made to follow the evolution of all seven families, but three of the best known, the 
dogs, bears, and cats, will be discussed. 

Second only to the higher primates in development of brain and intelligence 
are man's chosen companions, the dog and the cat; of the two, the useful dog ranks 
somewhat higher than the ornamental cat. There can be little doubt that the domesti- 
cated dog is a close ally of the modern wolf. In fact, the only obvious difference is 
that the pupil of the eye of the dog is circular, that of the wolf oblique. 

As compared with cats, modern wolves and dogs show relatively little departure 
from the structure of their Eocene ancestors. Almost the only carnivores which cap- 
ture prey by continuous pursuit, speed and stamina are their prime requisites. Speed 
has been improved through the ages by the lengthening of the legs, but this has not 
been accompanied by any great reduction in the number of toes. The small size of 
the ungual phalanges of the clawed animals makes it almost impossible to evolve a 
one- or two-toed foot (though to all intents and purposes this has been accomplished 
in the kangaroo). Moreover, although the carnivores pursue the hoofed mammals 
of the plains, their own habitations are commonly on the softer soils in the forests 
or along creek bottoms, where all toes come in contact with the irregular surface of 
the ground. Nevertheless, there is some reduction, for the thumb of the modern wolf 
or dog is vestigial, and there is no great toe. Functionally, therefore, the feet are 
four-toed. An artiodactyl-like feature is that the toes are in pairs of equal size, the 
inner somewhat larger than the outer ones; in this respect the feet suggest those of 
pigs. All dogs are digitigrade, with rather elongate feet, the weight borne on the pads 
under the distal phalanges. The claws are blunt, not retractile, of little use in the 
capture or dismembering of the prey but excellent for scratching on doors. The bones 
of the forearm show another adaptation to cursorial habits in that they have no power 
of rotation, the radius and ulna lying side by side. The skull is long, the brain large, 
with well-convoluted cerebrum. The dentition is primitive in many respects. Forty- 
two teeth are retained, the formula being i-j, cJ-, p-|, m-|-, The last upper molar is 
absent, and the third lower one is small, on the verge of disappearance. The upper 
molars retain to a remarkable extent the primitive triangular shape and tritubercular 
arrangement of cusps. 

Such are, in general, the characteristics of Cams familiaris, the dog, and Cants 


lupus, his uncivilized brother, the wolf. Dogs or wolves, henceforth in this section 
called dogs, were common during Pleistocene times. No representative of the direct 
ancestors is yet known from the Pliocene, but according to F. B. Loomis, whose out- 
line is followed here, Tomarctus, a small Mid-Miocene animal, differed but slightly 
from a true wolf. It is supposed to have been ancestral not only to Canis but also to 
Vulpes, the fox. Ancestral to Tomarctus is Nothocyon of the Lower Miocene, which 
is intermediate in size between a modern red fox and a coyote. Its dental formula 
is that of the dog, but the teeth are smaller and less widely separated. The thumb is 
much less reduced than that of modern canids, although distinctly shorter than the 
fingers. The foot has five digits of different lengths, for the fingers and toes are not 
in pairs as in the later forms. The claws are thin, sharp, somewhat retractile, and 
hence more catlike than those of dogs. It probably was not a particularly rapid runner. 
Fortunately, brain casts are known, and these show a smaller size and fewer con- 
volutions than those of modern wolves. 

In Oligocene strata are found skeletons of the somewhat smaller, foxlike Cynodic- 
tis (Fig. 118), with long body and tail, short, weak limbs, and five-toed feet, armed 
with sharp claws. The spreading wrist and ankle bones connote the large, loosely 
articulated feet of an animal not swift in pursuit but probably a stalker, awaiting the 
opportunity to make its kill. Incidentally, the big head and feet of your awkward 
puppy are souvenirs of his creodont ancestors. The dental formula is that of Canis, 
but the short face enforced a compact dentition, without diastemata. 

Although Cynodictis (now Pseudocynodictis} may be the first dog, its progenitor 
is found in the creodont family Miacidae (Fig. 119). There seems to be no doubt that 
the Eocene Miacis, the genus which has given its name to the family, included the 
ancestors of the dogs. Small in size, more like weasels than wolves, animals of this 
genus retained the typical forty-four teeth and had five functional toes on the short, 
spreading feet, each toe provided with a sharp, somewhat retractile claw. The brain 
was a little larger than that of other creodonts but only slightly convoluted. These 
creatures can hardly be called dogs, for they appear to have been ancestral not only 
to that family but to various other groups, some now extinct, others, such as the 
bears, still with us. 

Modern bears differ in striking fashion from dogs both in appearance and habits; 
yet they have the same ancestry. The skull is, it is true, rather doglike, although the 
width at the arches, reminiscent of the creodont ancestor, makes the face broad. 
Ursus, the bear, has the same number of teeth as Canis, but the anterior premolars, 
instead of being strong, shearing teeth, are small, three of them single-rooted, and 
of so little use that many individuals lose them early in life. Premolar - and molar - 
are not shearing but crushing teeth, entirely unlike true carnassials. The molars are 
longer than wide, with numerous tubercles on the crown, suggesting the pig rather 
than the carnivore. Like the pig and man, the bear is omnivorous. Those species 


which live where they can get it seem to prefer vegetable food, though they are not 
averse to such tidbits of flesh as come their way and are particularly keen for the nectar 
which bees have stored for their own use. Polar bears, on the other hand, are car- 
nivorous perforce, their diet largely piscine. Other characteristics of the bear are 
distinctly un-doglike. The body is short and heavy with a short tail, the limbs short 
and massive, the feet large, plantigrade, with five functional toes bearing long sharp 
claws. Unlike the dogs most of them climb trees readily, although, fortunately for 
man, the least amiable of their species, the grizzly, lacks this power. 

At the present day bears, although chiefly holarctic, are found on all continents 
except Australia. Apparently they were widespread during the Pleistocene also, some 
of the species then in existence belonging to the modern genus Ursus. During that 
period, however, the short-faced bear, Arctodus, or Arctotherium as some call it, 
ranged throughout North America from Pennsylvania, South Carolina, and California 
through Mexico, and penetrated far into South America. It was the only bear to 
reach the southern continent. It appears to have been ancestral to Ursus (Fig. 119). 
In Pliocene times bears were rare in North America, although Hyaenarctos, which 
was probably ancestral to Arctodus, did reach this continent. This genus appears to 
have been at home in southern Europe and Asia, where its remains have been found 
in France, Greece, and the Siwalik hills of India. It was more primitive than the 
modern bear in that only the first premolar was small and that the upper molars were 
quadrangular, rather than longer than wide. 

The ancestor of Hyaenarctos was French, not American, but this ancestor is 
classed by taxonomists not as a bear but as a dog, Hemicyon. Its remains are found 
in the Upper Miocene deposits at Sansan. It was digitigrade rather than plantigrade. 
The first upper molar was rounded-triangular rather than quadrate, and the second, 
oval and somewhat diagonal in position, as would be expected of a tooth modified 
from one of originally triangular form. 

According to W. D. Matthew, the Oligocene Daphoenus of America was the 
ancestor of Hemicyon. Several species are known, and paleontologists differ as to 
the interpretation of the fossils. They are classed as dogs, although they show many 
catlike characteristics. In spite of these features, they seem to have given rise to two 
lines, in one of which the teeth remained sharp and doglike, whereas in the other 
the bunodont condition of the bears was attained. This doglike line did not lead to 
modern dogs but produced huge, long-headed animals which died out during Plio- 
cene times. Not so perfectly adapted to their environment as the true dogs, they 
were unable to compete with them and so disappeared, whereas the omnivorous 
bearlike creatures, which competed with neither dogs nor cats, have persisted to the 
present. Daphoenus was small, not so large as a coyote. The teeth were small and 
closely set, the full number, forty-four, being present. Carnassials were present, not 
the sharp shearing blades of true sectorial teeth but forerunners of the blunt crushers 


of the bears, although they retained some of the shearing power of those of the 
Miacidae. The skull had a short face and relatively small brain case; the feet were 
spreading and digitigrade but did not show as great a departure from the plantigrade 
type as in the contemporaneous ancestor of the true dogs (Pseudocynodictis}. The 
claws were partially retractile. The fact that the lumbar vertebrae were large suggests 
powers of leaping, a catlike feature. All in all, this probable ancestor of the bears was 
little like the modern representatives of the group. The ancestor of Daphoenus was un- 
doubtedly some member of the creodont family Miacidae, probably Mtacis itself, 
ancestor, as we have seen, of the dogs. 

Why the antipathy between domestic cats and dogs? Is it because they have a 
common ancestry? Or is it, as owners of pets often suspect, merely a jealousy which 
has been engendered in the comparatively recent days in which both have been the 
associates of man? The latter explanation is the more probable, since the habits of 
cats and dogs are so unlike that, although both are carnivores, they are actual com- 
petitors in only a few habitats. The cat hides, lurks, crouches, and springs. Dogs 
openly give chase, generally in packs, their baying progress markedly in contrast 
to the silent, individualistic activities of the felines. These diverse activities bring 
into use different muscles; consequently the structures of the animals of the two 
groups have become increasingly unlike in the progress of time. In most respects 
cats are more specialized than dogs. Only in the quality of the brain does the dog 
excel all other carnivores. 

Cats have short, broad heads, which provide little room for the teeth. The denti- 
tion is correspondingly reduced, chiefly by the loss of most of the molars; typical 
cats have only one molar in each jaw, the upper one small, peglike, practically use- 
less, the lower a large sectorial which shears past the last upper premolar. The most 
specialized depend largely upon the last upper premolar and the last lower premolar 
and first lower molar for slicing their meat into pieces small enough to be swallowed; 
others have one more functional premolar of the shearing type on each side. But 
the shortening of the jaw is not the only cause of loss of teeth. Practically all felids 
have large canines, with which their prey is seized or killed. It is commonly the case 
that when canines or other teeth, for that matter are much enlarged the neighbor- 
ing ones become more or less functionless and tend to disappear. Thus, in cats, the 
first premolar is absent, the second is absent or vestigial, and the third, although 
generally present, at least in the upper jaw, may be small. 

The Felidae are divided into two subfamilies, the true cats (Felinae) and the 
saber-toothed tigers (Machairodontinae) . The dental formula of the former is 
i-1, c , p--^-, m~- X 2 = 28 30. The incisors are small, the canines large, oval 
in section, the lower nearly as large as the upper. Each jaw has two large premolars. 
The neck is short; the body and tail are long in most, although the tail of the lynx 


is short. The hand has five fingers, the thumb rather short; the foot has four toes, 
the first having been lost. 

Unfortunately, little is known of the genealogy of the true cats. They appear 
first in Lower Miocene deposits in both North America and Europe; thenceforth 
they seem to have been fairly common on both continents, but no direct lineage is 
traceable. The Lower Miocene Nimravus differs from the modern cat chiefly in 
having an additional small premolar in the lower jaw and somewhat more primitive 
carnassials. Curiously, the .upper canine was longer than the lower, laterally com- 
pressed, bladelike rather than oval in section, and hence comparable to the tusk of 
the saber-toothed tiger rather than to that of the cat. This suggests that the two 

FIG. 120. Skulls of three cats, showing the small amount of change during 
their known evolution. A, Dinictis, Oligocene; B, Nimravus, Miocene; and C, 
Felis y Recent. Redrawn from W. D. Matthew. Recent investigations throw 
doubt on Matthew's conclusions, expressed in the text, that Dinictis and 
Nimravus are ancestral to the Felinae. The former is probably a "saber- 
tooth," and the latter is considered the type of a third subfamily, intermediate 
between the two usually recognized. 

subfamilies of cats had a common ancestor, probably already identified by the ver- 
tebrate paleontologists in Dinictis, a primitive cat of the Oligocene. Although the 
latter has the long canines of the saber-tooths, it has many of the characteristics of the 
true cats. 

The saber-toothed tigers were the most spectacular and most highly specialized 
of the cats. As a subfamily they differed from the true cats in having bladelike 
canines, a flange on the lower jaw, five toes instead of four on the hind feet, a bony 
outgrowth on each ungual phalanx which served as a partial shield for the retractile 
claw, and a short tail. The history of these animals is as little known as that of the 
true cats. Like many other North American animals they achieved their greatest 
perfection and suffered extinction during the Pleistocene. Because of the abundance 
of its remains in the tar pits at Rancho La Brea, Smilodon, the last representative 
of this race, is commonly regarded as a Californian, perhaps a "native son." As a 
matter of fact, however, individuals belonging to various species of this genus ranged 


during Pleistocene times over the area from Pennsylvania and California in the 
north to the pampas of Argentina in the south. Larger and more terrible than any 
lion or tiger, imbued, seemingly, with a blood-lust for many paleontologists be- 
lieve that it killed for blood rather than for flesh Smilodon must have been the 
scourge of its age. In strength and ferocity it outranked the contemporaneous Ameri- 
can cat, Felis atrox, largest and strongest of all lions. Its terrible scimitar-like canines 
have gained for it the interest of all. These teeth, eight inches long in the largest 
individuals, are laterally compressed, with serrate posterior edges, perfect weapons 
for the severing of the jugular vein. The small size of the lower canines indicates 
that. the saber-tooth killed by striking, after the fashion of a snake, rather than by 

FIG. 121. Skulls of three saber-toothed tigers, showing the relatively 
small changes which have taken place during their evolution. A, Hoplopho- 
neus 9 Oligocene; B, Machairodus, Pliocene; C, Smilodon, Pleistocene. Re- 
drawn from W. D. Matthew. 

seizing with the jaws and rending with the claws, as true cats do. It may well be 
that this extraordinary development of the canines was what led to the extinction 
of the group, just as the overgrowth of the incisors of squirrels and other gnawing 
rodents locks the jaws of certain individuals and results in starvation. 

Little is known of American Pliocene or Miocene saber-tooths, although several 
species are known from our Oligocene. In Europe, however, several species of Machair- 
odus are found in deposits of Pliocene to Pleistocene age. These differ from Smilodon 
in that the body is smaller, the canines shorter, the flanges on the lower jaws larger, 
the brain case smaller, and the dentition less reduced. Smilodon has a dentition 
more reduced than that of any true cat; that is, [ -~^-, c-~, P-TTTTJ m-f. All species 

_ 3~"Z 1 ji ~~ 1 1 

of Machairodus have two lower premolars. The upper molar, although vestigial 
in both general, is behind the carnassial in the French genus, not internal to it as in 

Several species of saber-tooths are known from the Oligocene, some of them 


showing an admixture of truly feline characteristics. The Lower Oligocene Hoplo- 
phoneus appears to be ancestral to all later members of its race. The size was variable 
in the different species, the oldest being smaller than a modern wildcat. The simplest 
dental formula in the genus is i-~, c-j-, Py, m-j-, a total of 32. The carnassials are less 
specialized than those of other cats, in some respects doglike. The canines are long 
and bladelike, serrated on both edges, but do not project beyond the flanges of the 
lower jaw, these being deeper than in later forms. According to J. C. Merriam's 
interpretation, Hoplophoneus probably had habits exactly the opposite of those of 
the modern cats. He believes that it held its prey with its strongly clawed, grasping 
feet, while it struck repeatedly with its knifelike canines. 

A contemporary of Hoplophoneus, but somewhat more primitive, was Dinictis, 
an animal with the characteristics of the possible ancestor of all later Felidae. Its 
dental formula, i-i, c-j-, p-~, m^- (34 in all), was distinctly more primitive than that 
of any other cat. The upper molar, though diminutive, was of the tritubercular type. 
The skull was longer than that of the cats, the limbs were relatively longer and more 
slender, the five-toed feet small and weak. The claws were less retractile than those 
of other cats, and the gait almost plantigrade. It appears to be a connecting 
link between cats and dogs, and although its short tusks are laterally compressed 
it seems to be the prototype of both saber-toothed tigers and true Felinae. The 
views as to relationships expressed above are those of W. D. Matthew. Recent 
studies have thrown doubt upon some of his conclusions, but without setting up 
new lineages. 

Mammals seem to have been only moderately successful as terrestrial carnivores. 
An attempt has been made in the preceding pages to trace the lineages of a few of 
the more conspicuous groups. There are many others, but the creodonts, dogs, cats, 
hyenas, and bears are the ones that really deserve the title of beasts of prey. As has 
been seen, their evolution was parallel to that of the other mammals. On the whole 
they have been more variable, though less prolific in the production of species, than 
the herbivores. This statement seems contradictory and needs some elucidation. The 
point is that the terrestrial carnivores show but little of that program evolution which 
is so conspicuous a feature of the history of the hoofed quadrupeds. Horses, camels, 
rhinoceroses, tapirs, and other groups differentiated in the Eocene or earlier, and 
each pursued its own path. There was wide variation within narrow limits. Among 
the beasts of prey, on the other hand, there is no clear-cut lineage extending back to 
the Eocene. The dogs furnish the nearest approach to one, but an Eocene dog was not 
a dog in the same sense that an Eocene horse was a horse. The creodonts held the stage 
during the first half of the Tertiary, variation producing among them doglike, catlike, 
hyenalike, bearlike, and other sorts of creatures. With their decay in the Oligocene, 
the Fissipedia emerged and through a similar series of variations produced the modern 


carnivores. But it was not till Miocene times that dogs, cats, and bears were fully 
differentiated. These animals reached their maximum in the Pliocene and Pleisto- 
cene. Their rapid downfall has been due to such a wholly unpredictable combination 
of circumstances that they can hardly be blamed for having mistaken their autumn 
for their prime. 

FIG. 122. Restoration of the Pleistocene saber-toothed tiger, redrawn after 
a painting by Charles R. Knight. 


No man can tell what the future may bring forth, and small opportunities are often 
the beginning of great enterprises. Demosthenes, Ad Leptinem 

The formation of great plains high above sea level east of the Cordilleran-Andean 
ranges of the Americas and north of the Himalayas in Asia was a result of tremendous 
earth movements at various times during the Tertiary. None too well watered, the 
vegetation of these plains gradually changed; forests became thinner and thinner, 
giving opportunity for the spread of grasses and shrubs, till then held in check by 
the shade of great trees. Finally a condition was reached in which trees and lush 
herbage were restricted to relatively small areas in river bottoms, whereas extensive 
regions were occupied by grasses. These were tender in the spring of the year, but 
as the season advanced they became tough and wiry. Dry grasses have high food 
value; even in winter they serve to sustain life, though animals may have to seek 
them beneath a cover of snow. Food of any sort is always at a premium; hence various 
mammals left the forests and invaded the ever-increasing areas of the plains. Some 
may have done so from choice, others by accident, but environmental pressure was 
probably the chief motivating cause. 

As the restriction of the forests reduced the amount of food in them, their area 
became too small to support the growing population. It is the same sort of pressure 
which is troubling the world today, that pressure of overpopulation which causes 
wars, sometimes of physical combat, sometimes battles of wits called diplomacy. Just 
as human beings are forced by economic necessity to seek homes in unaccustomed 
areas where they must learn to eat new foods and protect themselves from new 
enemies, so were the Tertiary mammals compelled to adapt themselves to new cir- 
cumstances. Fortunately for the success of mammalian evolution, they did not suffer 
the sudden transitions which have fallen to the lot of man. The Englishman or the 
German, transplanted to equatorial Africa or the islands of the South Seas, tries to 
take his environment with him, and is more or less successful. It is largely this which 
enables him to "carry on." Mammals with less brain could not do so, nor was it 
necessary, for climates change slowly. Young horse or deer or antelope never re- 
marked, "Leaves are scarcer here than they were in great-grandfather's time. We'll 
simply have to eat grass." Mammals had to eat, and they ate what they could get. 
Times were hard, undoubtedly. There was a worldwide depression. Many a rich 
and noble family went to the wall. But those which learned to eat grass found a 


new source of income. "Natural resources" were not by any means exhausted. On 
the basis of grass new family fortunes were founded. 

Carnivores could not flourish in this new environment, in which there were few 
opportunities for concealment. Decrease in number of enemies allowed greater ex- 
pansion of the herbivores. So in the Mid-Tertiary a new dynasty arose, that of the 
hoofed animals, the Ungulata, whose reign was checked only in part by the glacial 
climates of the Pleistocene but has been brought almost to a close by the prowess of 
man, first in hunting and later in agriculture. 

The first hoofed animals lived in the Eocene forests, feeding on leaves, coarse 
vegetation, and probably to some extent on such animal food as happened to be 
available insects, worms, snails, perhaps occasionally a bird, a lizard, or a small 
mammal. Their teeth, as has already been pointed out, did not differ greatly from 
those of contemporary flesh-eaters. Primitive ungulates differed only slightly from 
primitive carnivores, and there are indications that the former group is a branch of 
the latter. The chief specializations of the hoofed quadrupeds have been the lengthen- 
ing of the legs, the reduction in the number of toes as the animals became more and 
more adapted to running on the open prairies, and the evolution of grinding crowns 
on the molar teeth, a modification which enabled them to subsist on tough dried 
grasses. Some races of hoofed animals, however, continued throughout their history 
to feed upon coarse vegetation, which can be obtained on the plains as well as in the 
wooded areas; others remained in the forests. All such browsers have retained a 
rather primitive dentition, teeth better adapted for crushing than for grinding. 

The Ungulata are readily subdivided into two great orders, the Perissodactyla 
and the Artiodactyla. Numerous species belonging to both orders still exist; even 
more diversified are the extinct forms. North America has been the principal home 
of the perissodactyls throughout their history; Eurasia and Africa the locus of the 
evolution of the artiodactyls. There are notable exceptions to this generalization, for 
the Siberian-Alaskan land-bridge repeatedly permitted the interchange of faunas 
during Tertiary times. 

There are several marked differences between the two orders. The Perissodactyla 
are the odd-toed hoofed mammals, the horse and the rhinoceros being the most familiar 
modern examples. Although the modern horse has one toe and the modern rhinoceros 
three, the numbers obviously odd, the tapir has four on the front feet, and various 
extinct perissodactyls had an even number of toes. What is really meant when one 
speaks of an odd-toed animal is that the foot is mesaxonic; that is, the middle toe 
is the largest, the axial plane passing through it (Fig. 123 B). Hence this digit is 
bilaterally symmetrical. Another outstanding characteristic of the Perissodactyla is 
that in all existing animals of the order all the premolars except the first resemble 
the molars. This modification came about gradually, for few of their ancestors have 
teeth so constructed. One of the upper bones of the ankle, the astragalus, is also useful 



in distinguishing the two groups. That of a perissodactyl is deeply grooved above 
(Fig. 123 B, C), where it articulates with the principal lower bone (tibia) of the 
leg, but flat beneath; the astragalus of the artiodactyl (Fig. 123 D) is grooved on both 
upper and lower surfaces. There are, of course, many other characteristics of the 
skeleton which afford the initiates in vertebrate paleontology instant information 
but which have little significance to the novice. 

Since the history of the perissodactyls is best documented in North America, it 
follows that they have been intensively studied by American paleontologists, who 
have described in detail the ancestry of the horses and the titanotheres and have made 
considerable progress in tracing the various families of rhinoceroses. The discussion 

B c 

FIG. 123. A, fore foot of the camel to illustrate the typical artiodactyl 
structure. After W. H. Flower. B, hind foot of Hyrachyus, an Eocene rhinoc- 
eroid, as an example of a perissodactyl foot. After E. D. Cope. C, two views 
of the hind foot of Moropus, the "clawed" perissodactyl. After O. A. Peter- 
son. D, the doubly grooved astragalus characteristic of the artiodactyls. 

of two of these groups is undertaken in a later chapter entitled "Some Genealogies." 
This leaves only the tapirs and the strange, aberrant ancylopods to be described here. 

Because of its many primitive characteristics, the tapir has repeatedly been called 
a "living fossil" (Fig. 125) . Nevertheless, it cannot be considered an unchanged sur- 
vivor of the Eocene fauna. No race of vertebrates, so far as is known, has survived 
any really long period of time without some change. All that is implied in the term 
quoted above is that the specializations are relatively inconspicuous. 

Modern tapirs are more specialized than their ancestors in that they are larger, 
although none has reached elephantine or even rhinocerine proportions. They are 
primitive in that they retain the typical forty-four teeth. But the teeth themselves 
have been somewhat specialized. The third upper incisor has become so large as to 
take over the function of the canine, the true canine is small, and there is a long 
diastema between the canine and the first premolar; the premolars have much the 
same structure as the molars; the latter, however, are simple, for they are quadrangular, 
each bearing but one pair of cross crests, obviously no great modification of the 
primitive cusps of the tritubercular tooth. The nose and upper lip of the modern 

FIG. 124. Teeth of Lophiodon, a French Eocene tapiroid, as an example 
of brachydont teeth with lophodont pattern in the arrangement of the cusps. 
Natural size. From F. L. Pictet, Traitc de paleontologie. 

FIG. 125. At left, the skull of Miotapirus, a Lower Miocene tapir. At 
right, drawings of the left upper cheek teeth, to show the evolution from 
the Eocene to the present. A, Hcptodon, Eocene; B, C, D, three species of 
Protapirus from successive zones in the Oligocene; E, Miotapirus, Lower Mio- 
cene; F, Tapirus. From Eric Schlaikjer. 

FIG. 126. Teeth of a true pig, Choeromorus, from the Upper Eocene of 
France. Natural size. From F. L. Pictet, Traite de paleontologie. 

FIG. 127. The Lower Miocene Dinohyus, the largest of the entelodonts. 
From O. A. Peterson. 

FIG. 128. Merycoidodon, from a specimen, seventeen inches high at the 
shoulder, in the Museum of Comparative Zoology, Harvard University. 


tapir are combined in a short proboscis, reflected in the skull by short frontal bones 
and the separation of the nasals from the maxillaries and the premaxillaries; but 
this specialization is hardly comparable to that of the elephants. The legs are slender 
but not long. The front feet have four, the hind three short toes, in no way com- 
parable to those of the horse. On the whole the primitive outweigh the specialized 

The geographical distribution of modern tapirs is that which is to be expected 
of a primitive group of mammals. As Matthew demonstrated, many archaic crea- 
tures have survived till the present day because, as climatic conditions became in- 
creasingly rigorous in the regions which they commonly inhabited, they retreated, 
somewhat ingloriously, to the tropics, where food is so abundant as to enable even 
the most backward to survive. Tapirs illustrate this admirably, for they are found 
now only in Central and South America, in southern Asia, and in the adjacent 
islands. Such a distribution is considered to indicate the great antiquity of the race, 
for there has been no possibility of direct migration of terrestrial animals from tropi- 
cal America to tropical Asia within the period covered by the known history of 

Because tapirs have always been forest animals, and because the fresh-water 
formations from which most Tertiary mammalian fossils are obtained were de- 
posited in flood plains and lakes on the prairies rather than in the forests, little is 
known of the lineage of the family. During the Pleistocene, representatives of the 
modern genus Tapirus appear to have been widespread, particularly in the northern 
hemisphere, their remains having been found in the United States, Europe, and 
China. A few penetrated to South America at this time, but none reached Africa 
or Australia. 

Little is known of Pliocene or Miocene tapirs (Fig. 125) ; just enough scraps and 
fragments of bone have come to light to show that they were present in the northern 
hemisphere during the later part of the Tertiary. The older Oligocene beds of 
North America and the Mid-Oligocene of Europe have, however, produced skulls 
and skeletal material pertaining to a genus which has been aptly named Protapirus, 
for it is obviously a primitive tapir. Only about half the size of modern members of the 
family, it nevertheless foreshadows those now existing, for its molars are of the same 
pattern and the position of the nasals shows that a short proboscis was present. If 
observed in greater detail, however, it will be noted that the true canines were 
functional, since the outer incisors are not enlarged. Furthermore, the premolars 
are not like the molars but have a distinctive pattern, retaining some resemblance to 
the primitive triangular shape. In these respects Protapirus is more primitive than 
Tapirus, but there is no such striking change in appearance and structure as there is 
between Oligocene and Pleistocene horses. 

Although the history of the tapirs during the Eocene is not yet supported by 


full information, its general outline can be grasped. The earliest Eocene tapirs, 
like the contemporary horses, were inhabitants of the forests. Both were small ani- 
mals; both fed on coarse vegetation which was crushed rather than chewed. As con- 
ditions changed, the more adventurous horses emigrated to the grass-bearing plains, 
where their teeth and legs were more and more modified in such ways as to adapt 
them to successful life in the new environment. The more timid tapirs, on the other 
hand, risking less, clung to the wooded areas. From an evolutionary standpoint it is 
interesting to note that even though they remained in their original environment so 
far as possible they could not avoid sharing in some measure the effects of the 
general trend of changes undergone by most mammals during the Cenozoic. In the 
course of time all members of the family increased in size; the brain became larger; 
there was some, although not great, reduction in the number of toes. Most striking 
of all changes, the premolars gradually became molariform. This last cannot be 
ascribed to change in diet, for it is a characteristic of odd-toed mammals as dis- 
tinguished from those whose digits are paired. 

Can it be inferred from this that there are two sets of factors which cause changes 
in animals, one external, the other internal? Increase in size and sagacity and de- 
crease in number of toes can all be ascribed to adjustment to the habitat, although 
terrestrial environments are not everywhere the same. Throughout the Cenozoic 
there was constant change of temperature, of kind and amount of vegetation, and 
of degree of competition for food. Reaction to none of these changes, however, will 
explain why the premolars of perissodactyls of every kind plains-dwelling horses, 
woods-loving tapirs, cursorial, semiaquatic, and giraffe-like rhinoceroses exhibit the 
same tendency to become like the molars. That such a change appears in all lines 
suggests that it is a matter of heredity. Was there something in the "blood" (chromo- 
somes) of the ancestor which came to light sooner or later in all the offspring? 

The tapirs are the most primitive of the perissodactyls, the horses the most spe- 
cialized, but the chalicotheres are the most curious. They are classified as hoofed 
mammals; yet they have claws and no hoofs. They are called perissodactyls; yet the 
second, not the third toe is the largest (Fig. 123 C). On the other hand, although 
they have claws, they are not carnivores, for the teeth are of the typical herbivorous 
type, much like those of the titanotheres to which some paleontologists believe they 
are related. Most of them have a foot in which the second toe is the largest, but there 
is evidence that their ancestors were mesaxonic, as the other perissodactyls are. The 
chalicotheres are commonly cited as the great exception to Cuvier's "law" of the cor- 
relation of parts. When the first fragmentary specimens were found, the skull was 
interpreted as that of a perissodactyl and the feet as those of a pangolin (scaly ant- 
eater), the parts being thus distributed in two unlike orders. The large, deeply cleft 
ungual phalanges are not, however, exactly duplicated among the known carnivores. 

Remains are mostly fragmentary and decidedly scarce; yet chalicotheres are 


known to have had a long history in the northern hemisphere. The oldest repre- 
sentatives are found in the North American Eocene; they occupied this country till 
the Miocene; in Europe and Asia they existed from the Oligocene to the Pliocene; they 
may have reached Africa in the Pliocene, but the record is doubtful. The best-known 
American ancylopod is the early Miocene Moropus, nearly complete skeletons of 
which are to be seen in the museums in Pittsburgh, New York, and Cambridge. 
It is a large animal, somewhat giraffe-like in aspect, the tore legs longer than the 
hind, and the neck elongate. The skull is small, the body long, sloping from shoulder 
to rump. Each foot had but three functional toes; nothing remains of the first, but 
the fifth is represented by a small, vestigial metacarpal in the manus, although it is 
entirely absent from the pes. The fore feet, which are larger than the hind, are some- 
what perissodactyl-like in that the middle toe is the longest, although the second is 
stronger, with a larger claw. The European Chalicothcrium^ of Mid-Miocene and 
Pliocene age, was larger than Moropus but had a shorter neck. The Oligocene and 
Eocene members of the group were of smaller size. 

Although the chalicotheres were so giraffe-like in general appearance, there is 
no evidence of any close relationship to that animal. The chalicotheres, the ancestral 
giraffes, and the huge rhinoceros, Baluchitherium, appear to have had similar propen- 
sities for gathering their food from the branches of trees. Of the three, only the ex- 
tremely long-necked giraffe succeeded in changing to a diet of grass. And an 
exceedingly awkward job he makes of feeding, if we are correctly informed by the 
motion pictures. Nevertheless he has survived, whereas the others have not. 

The artiodactyls of the present day are much more numerous and much more 
diversified than the perissodactyls. Pigs, cattle, and deer are familiar, but all of the 
first and the domesticated members of the second of these groups were brought to 
this continent by man. North America has had but few of the cloven-hoofed beasts 
as compared with Asia and Africa, where they now exist in great numbers and 
variety. This is unfortunate, for occurrences of fossiliferous Tertiary deposits of any 
great extent are rare in the Old World. A few localities in France, Greece, India, 
Mongolia, and China are the only ones in the northern hemisphere which have 
yielded many specimens. The record for Africa is even less satisfactory, for aside 
from the remarkable Eocene and Oligocene deposits in the Fayum desert of Egypt, 
the continent has produced almost nothing older than the Pliocene. Because of this 
dearth of material it has not been possible to trace artiodactyl lineages in any great 
detail, except for that of the camel, which belongs to one of the two groups which are 
North America's chief contribution to the even-toed mammals. 

The general trend in artiodactyl evolution is the same as in that of the perisso- 
dactyls. The teeth become more complicated as the change is made from a diet of 
tender herbage and leaves to one of dry grasses. The body size increases, the cerebrum 
becomes more convoluted, and the legs are longer. The lengthening of the legs is 


accompanied by loss of one or more toes. Since the weight is carried chiefly on the 
third and fourth toes, the foot becomes symmetrical with regard to a plane passing 
between them (paraxonic) (Fig. 123 A). The first toe is vestigial or is lost; the second 
and fifth are in most cases smaller than the third and fourth. In all the advanced 
groups, the really rapid runners, the lateral toes are reduced to mere vestiges, the 
radius and ulna, and tibia and fibula are united, and the two functional metapodials 
are fused, producing the cannon bone. It is this condition, in which two toes are 
articulated to one bone, which gave rise to the appellation "cloven hoof." This ex- 
pression has too often been interpreted as meaning that a single hoof has been divided 
into two, whereas in reality the upper bones of two toes have coalesced. The other 
striking peculiarity of the artiodactyls, the doubly-grooved astragalus, has already been 

Modern artiodactyls fall readily into natural groups, but paleontologists have ex- 
perienced great difficulty in formulating a classification which will include all the 
fossil forms. In general, two great series may be recognized, one composed of the 
relatively primitive creatures which have a bunodont dentition, and another of a 
more advanced type with selenodont molars. The bunodont tooth (Fig. 126), short- 
crowned, with blunt conical cusps, like that of the pig, is well adapted for crushing 
any sort of food, hard or soft, animal or vegetable. Although they are primarily 
vegetarians, artiodactyls with such teeth are really omnivorous; most of them retain 
some characteristics of the creodonts. The selenodont molar (Fig. 129), short-crowned 
(brachydont) in its earlier manifestations but high-crowned (hypsodont) when per- 
fected, is the common possession of cattle, deer, giraffes, camels, and other artiodac- 
tyls. This tooth is remarkable in that the crests are longitudinal instead of transverse, 
as is the condition in most grinding teeth. Each molar has two pairs of lunate "lakes" 
of dentine, bordered by thin sharp rims of enamel. In some groups there are small 
modifications of this simple plan, but in general all selenodont molars are so much 
alike that they are of comparatively little use in the identification of animals. For- 
tunately, there are few instances among the artiodactyls in which the premolars be- 
come molariform, and F. B. Loomis has shown that the structure of the last premolar 
is an important guide to the recognition of the various subdivisions of the group. 

The Old World pigs and the New World peccaries are the best examples of the 
bunodonts. Although superficially much alike, they are assigned to separate families 
which, when united with the Hippopotamidae, make up the superfamily of the 
Suioidea, piglike creatures. 

The Old World swine are the most primitive, for the typical dentition of forty- 
four teeth is retained and each foot has four toes, although the lateral ones are func- 
tional only on marshy ground. It is well known that the pig will eat anything, the 
tubercle-studded molars crushing nuts and bones as readily as they do the refuse on 
which the domestic animal is commonly fed. The records of boar hunts sufficiently 


attest the ferocity and agility of the carnivore-like creatures, armed with triangular 
canines of the most dangerous sort. Once widespread throughout the forested regions 
of Europe, northern Africa, and central Asia, Sus scrofa was one of the great beasts 
of the chase during the Middle Ages. The pursuit of the boar was the sport of kings; 
hence, by reflected glory, the animal became royal, and one of the four used in heraldic 
devites. Richard the Third, "Crookback," murderer of the Princes, employed it on 
his shield. Remains of pigs found in strata of various ages from the Oligocene on- 
ward show that they have long been inhabitants of Eurasia and Africa, but they never 
migrated to North America. Since the living creatures are so primitive, it is not 
surprising that the fossils show no spectacular changes during the evolution of the 

Peccaries, the New World swine, have much the same appearance as the true 
pigs, although the absence of a tail would probably be noted by a sharp-eyed small 
boy. In reality, however, the peccaries are somewhat specialized. Instead of forty-four 
teeth, there are only thirty-eight, the third upper incisors and the first premolars 
having been lost. The upper canines point directly downward, not outward and 
upward as in the pigs. The molars are quadrangular, with four principal cusps, to 
which small wartlike tubercles are added, whereas in the European swine the cheek 
teeth are longer than wide, the chief cusps irregularly arranged. The last lower 
premolar of the peccary is molariform, an unusual condition for an artiodactyl. The 
fore foot has as many toes as the pig, but the hind has only two, for the second is 
vestigial and the first and fifth are absent. The metapodials are fused at their upper 
ends, forming a rudimentary cannon bone. At the present day peccaries are chiefly 
South American, although one species is found throughout Mexico and as far north 
as Texas and New Mexico. They are, however, primarily North American animals, 
for they have been here since the Oligocene and did not enter the southern continent 
until the Pleistocene. The Eocene ancestors of true pigs and peccaries are still unidenti- 
fied. It will probably be shown eventually that they arose from the same stock, either 
in Eurasia or in North America. 

Another family of bunodonts which flourished in North America during the 
Oligocene and Miocene (remains are also known from the European and Asiatic 
Oligocene) is that known as the entelodonts (Fig. 127). Although they left no de- 
scendants, they are of interest because of the great similarity of their teeth to those of 
carnivores. The incisors are conical, the canines powerful tusks, and the premolars 
trenchant, reminding one of those of the dog, but the molars are small and piglike. 
Despite this savage dentition, specimens in the Princeton museum seem to show that 
the entelodonts fed principally upon roots, for the canines and incisors are scored 
with grooves which could have been produced only by such sand-covered food. 

The last of this line was Dinohyus, the largest of bunodonts other than the 
modern hippopotamus. This piglike giant, six feet high at the shoulders, was dis- 


covered by Olaf A. Peterson in the Lower Miocene strata of northwestern Nebraska, 
and its skeleton is mounted in the Carnegie Museum. Its most striking feature is 
the enormous skull, three feet in length, armed with splendid teeth. Long dorsal 
spines, for the attachment of the muscles necessary to support so large a head, project 
above the dorsal vertebrae, forming the nucleus of a hump at the shoulders. The legs 
are rather slender, the ulna and fibula united with their corresponding bones, the 
toes reduced to two. Curiously, no cannon bone was formed. This "terrible hog" 
was certainly not the ancestor of modern pigs or peccaries, for its feet were more 
specialized than those of either group. 

The entelodonts present a curious mixture of characteristics to those who like 
to speculate upon the probable habits of extinct animals. One would judge from 
the structure of the feet that they were accustomed to life on the plains; yet they 
disappeared just at the time when the prairies were in their period of greatest ex- 
pansion. If they were root-eaters, one would suppose that a forest would have been 
their natural habitat, particularly a moist forest. Perhaps they were harvesters of 
wild turnips, carrots, and other tuberous roots, although these could not have been 
easily dug by a hoofed mammal. 

Every collector of vertebrate fossils and every curator of a paleontological col- 
lection knows that the "oreodons" were the most numerous of all North American 
Tertiary mammals. Merycoidodon (Fig. 128), to give the genus the name con- 
sidered correct at the present time, seems to have been a gregarious inhabitant of the 
Great Plains during the Oligocene. Not only do the merycoidodonts appear to have 
existed in great herds, but individuals seem to have been almost unduly prone to 
early death and rapid burial. Had quicksands a morbid fascination for them? It 
would be interesting to know how many skulls of Merycoidodon culbertsoni have 
been collected. Dealers have handled them for years; they evoke endless sniffs from 
curators as collectors unpack the summer's spoils. Yet it is doubtful if enough material 
has yet been amassed to allow a really thorough study of this one species. Museums 
hesitate to discard specimens, no matter how many duplicates they may accumulate. 
Scientific inquiry is so many-sided that no one can foretell the moment when old 
material may be studied from a new standpoint. Thus, when M. R. Thorpe, trying 
to ascertain the reason for their extinction, studied "several hundred specimens of 
the Great Plains Merycoidodon and the John Day Eporeodon for the purpose of 
determining the relative age of the individuals at death," he would probably have 
reached more satisfactory results if he had had several thousand. 

The merycoidodonts are among the simplest of the selenodont (Fig. 129) mam- 
mals and, as a group, one of the most successful, for they appeared first in late 
Eocene times and survived until the Mid-Pliocene. Why they were extinguished at 
that particular time is a mystery, for it was not a critical period for mammals. Some 
thirty genera and more than a hundred species of this family are known. 


No merycoidodont was large or highly specialized. The form of mos' was 
piglike, though rather longer in the neck; the legs were short, with four toes on 
each foot. There was no real relationship to the pigs, however, for the cheek teeth 
are typically selenodont. A feature which at first glance seems peculiar is that these 
animals appear to have four lower incisors on each side of the symphysis. This is 
due to the fact that the first lower premolar has the shape and function of a canine, 
whereas the true canine is small, like an incisor. The result of this change is that the 
lower "canine" closes behind the upper instead of in front of it as in most mammals. 
This is not, however, a condition restricted to the family, for all ruminants except the 
camels show the same arrangement. The dentition is not that of a grazing animal, 
all the premolars and molars being sharp cutting and crushing teeth. Thus, although 
their remains are found associated with those of true ruminants, it must be inferred 
that the merycoidodonts dwelt by the water courses, living largely on leaves, young 

FIG. 129. The merycoidodont, Merychyus, to illustrate the selenodont 
pattern of the molar teeth. One-fourth natural size. From O. A. Peterson. 

twigs, and succulent vegetation. They seem to have been ill adapted to withstand 
winters or life away from areas of permanent water. This may explain both the 
final extinction of the group and their relative abundance as fossils. 

The Eocene merycoidodonts were small, but their dentition is so much like that 
of later ones as to give no clue to their origin. They are supposed to have been immi- 
grants to North America during the late Eocene. There is, however, no suggestion 
as to the locality of the original home. It is not unlikely that ancestors will eventually 
be found in our older Eocene formations. 

The group became considerably differentiated and widespread in North America 
during the Oligocene. The central stock seems to have been a cursorial type from 
which were derived some forest dwellers, some creatures of semiaquatic habits, and, 
most specialized of all, a few browsers with a tapir-like proboscis. The largest mery- 
coidodont, Promerycochoerus, lived during late Oligocene and early Miocene times 
Individuals were as large as modern hogs, though they can hardly have been so fat, 
One of the most remarkable groups of fossils ever collected consists of skeletons ol 
a species of this group, obtained by Peterson in western Nebraska, and now in the 
Carnegie Museum in Pittsburgh. On a single slab, lying in their natural positions. 


are three individuals, remains of animals which had crept close to one another for 

mutual warmth or consolation at a time when some disaster overwhelmed them. 

They may have succumbed to the cold of a western blizzard, but more probably they 

were the victims of a sandstorm which suffocated and buried them. A similar group 

of merycoidodonts may be seen in the American Museum of Natural History in New 


Although there are no definite connecting links, it may be confidently asserted 
that the hoofed mammals were derived from some as yet unknown carnivore of the 
creodont family. This relationship appears to find more obvious expression among 
the artiodactyls than among the perissodactyls. It has already been pointed out that 
the dentition of the entelodonts was strikingly carnivore-like. Thorpe has listed 
many characteristics of the merycoidodonts which attest their relationship to primitive 
flesh-eaters. As early as 1896 Joseph Leidy, the Philadelphia physician who brought 
this family into the scientific world, commented on the "decidedly wolfish aspect" of 
the skull, as indicated by the elongate form and almost continuous series of teeth. 
Among other characteristics, to which attention has been drawn by Thorpe, the 
most obvious are that the head is large, the face is long, as is the area in which the 
lower jaws are joined, the auditory bullae are not ossified, a common condition in 
the early creodonts, and the zygomatic arches are stout and heavy. So much for the 
skull. The body was long, the legs short, the tail long all characteristics of creo- 
donts rather than of ungulates. In addition Thorpe lists various characteristics of the 
bones of the fore and hind legs which are significant to the osteologist but of too 
technical a nature to be repeated here. The evidence is cumulative in support of the 
theory that clawed and hoofed mammals had a common ancestor. 

All in all, the history of the merycoidodonts is comparable to that of successful 
and contented burghers rather than to that of noble families. They lived long, with- 
out much change; they produced no giants, but no pygmies; they were numerous 
without being important; we learn nothing in particular from their history; their 
coming marked no epoch; they were not missed at their departure. How like the 
average citizen! 

Most important, most numerous, and most specialized of the artiodactyls are the 
ruminants, the cud-chewing grazers so important a source of food and clothing for 
man. Without the docile cattle, the stupid sheep, the hardy goats, and the. rugged 
reindeer, man, primitive or civilized, could hardly exist. Members of this group, 
known technically as the Pecora, lack upper incisors, the tongue performing their 
functions. Probably most readers have seen the cow gather wisps of grass, twist 
them against the lower incisors, and pull. This process is different from the cropping 
bite of the horse, but fully as effective, as sheep raisers know. Few ruminants have 
upper canines; the corresponding lower teeth are small, incisor-like; all lack the first 
premolar. The molars are selenodont, less cuspate than those of the mervcoidodonts 


The legs are long and slender, perfectly adapted for rapid motion. Some of the 
antelopes are, indeed, the fleetest of living animals. Although there are two toes, 
the leg is practically a chain of single bones, for the ulna is coossified with the radius, 
the fibula practically absent, and the two complete metapodials united as a cannon 
bone. It is noteworthy that the terminal processes, which articulate with the phalanges, 
are parallel; hence the toes are close together, not divergent, as are those of the some- 
what splay-footed camel. "Dewclaws," a pair of vestigial fingers and toes (second 
and fifth), are present on many Pecora but fail to impede the action of the axial 
portion of the leg, for their short metapodials do not in most cases articulate with the 
bones of either wrist or ankle, nor are they attached to the cannon bone. Although 
apparently excess baggage, they are not large enough to be of any great hindrance in 
running and may be of some use in traversing marshy ground. 

Males of most of the Pecora have horns, the few exceptions occurring amongst 
the deer. In the absence of any marked differences in the teeth, the horns have been 
used largely in classification. Although not at the moment in favor, the old terms 
Cervicornia (deerlike, solid-horned) and Cavicornia (hollow-horned) serve well to 
distinguish the two great groups of ruminants. The Cervicornia include the deer 
and giraffes, the former being much the more important. 

The antlers of the deer are bony outgrowths formed and shed annually. During 
the period of growth they are said to be "in the velvet," for they are then covered 
with a layer of skin which is shed when the horns are complete. After the rutting 
season a part of the bone just below the "burr" near the proximal end of the antler 
is resorbed, and the greater part of the structure falls off, leaving only a short per- 
manent projection of the frontals, known as the pedicle. Some deer have simple, 
undivided antlers; others carry magnificent but burdensome structures, many- 
branched^ with the numerous tines coveted by sportsmen. 

The oldest deer now known are small creatures without antlers, described from 
European Oligocene deposits. This simple type is apparently represented today by 
the Chinese water deer, which, like other deer without antlers, or with small ones, 
has a pair of long sharp recurved upper canines, effective weapons. In this line the 
primitive canine has been increasingly developed throughout the successive genera- 
tions. Lower Miocene strata, again in Europe, have produced the earliest deer with 
antlers, forms referred to Palaeomeryx. Animals belonging to this genus have a 
pair of simple bony processes, probably covered with skin, over the orbits. From 
such ancestors, according to Loomis, who has studied the evidence in detail, were 
derived on the one hand the deer with deciduous antlers and on the other the 
giraffes. In this latter group the horns are not shed but are skin-covered bony 
outgrowths, comparable to the pedicles of the deer. 

Evidence obtained from European fossils shows that the evolution of true deer 
with antlers was rapid after their first appearance in the early Miocene. A few strag- 


gled to America at this time but seem not to have prospered. Our oldest is the hornless 
Blastomeryx, whose remains have been found in strata of various ages from Lower 
Miocene to Pliocene. It was a small, lightly built creature, with saberlike canines 
and legs about as specialized as those of modern deer. It seems to have been ancestral 
to Merycodus, species of which existed in North America from Mid-Miocene to early 
Pleistocene. Merycodus had simple, two- or three-pronged antlers but was probably 
not ancestral to any modern deer, for its teeth were more fully hypsodont than are 
those of any living member of the family. The earliest Merycodus was only eighteen 
to twenty inches high at the shoulder; yet the legs were long and slender, without 
traces of the lateral toes. Thus both teeth and legs were more specialized than those 
of modern deer. 

It was not until late Pliocene or early Pleistocene times that deer became really 
plentiful in North America. All that have left descendants appear to have come from 
Asia, although there is a possibility that some of the more southern forms may trace 
to some as yet unknown American Miocene animal. According to W. B. Scott, there 
were two immigrations, the first of which, in the Pliocene, brought in the ancestors 
of the Virginia deer and its allies. These animals spread not only over all North 
America but into the southern continent, where they multiplied and became con- 
siderably diversified. From an evolutionary standpoint it is interesting to note that, 
since they were the only South American ruminants, they found no great competition 
from animals with their own food habits when they arrived in Pliocene or early 
Pleistocene times. The second invasion of North America was in the Pleistocene 
and was marked by the immigration of deer which have remained northern in their 
distribution. Among them were the reindeer, the palmate-horned moose, and the 
many-pronged Wapiti, all animals which strayed into what is now the southern 
United States only during periods of glacial advance. 

The history of the giraffes, which, as has been seen, appears to have begun in 
the Lower Miocene with Palaeomeryx, is obscure at the moment. The living mem- 
bers of the family, the giraffes and the rare okapi, are confined to the part of Africa 
south of the Sahara, but remains of extinct species have been found in deposits of 
early Pliocene age in Hungary, Greece, Persia, northern India, China, and East 
Africa. The modern giraffe is notable for its elongate neck and strong forelegs. The 
smaller . okapi, with a much shorter neck and approximately equal fore and hind 
legs, appears to be much more like its Pliocene forebears than the giraffe is. This 
seems to indicate that the most striking specializations in this line have been acquired 
rather recently. European records older than the Pliocene are strangely silent as to 
the ancestry of the giraffes; hence it is probable that Loomis is right in asserting 
that the American Miocene Dromomeryx is the real connecting link between this 
group and the Lower Miocene ancestor of the antlered deer, Palaeomeryx. Dromo- 
meryx (Fig. 130) is a small, Mid-Miocene artiodactyl with a pair of simple, skin- 


covered horns above the eyes. That it has been considered the oldest American 
antelope by one student of fossil mammals and the oldest giraffe by another is of 
interest to the layman chiefly as evidence of the affinity of the Cervicornia to the 

The latter are the most specialized of the ruminants, best fitted for life on the 
prairies. Their hypsodont teeth are adapted to a diet of grasses, fresh or dry, whereas 
the short-crowned teeth of most of the Cervicornia indicate that they were primarily 
browsers. The term "hollow-horned" refers only to the horny sheaths which cover 
the bony outgrowths from the frontal bones, present in females as well as males, ex- 
cept .for the does of some species of antelopes. Neither bony core nor horny sheath is 

FIG. 130. A simple-horned Miocene deer, Dromomeryx. One twenty- 
fourth natural size. From E. Douglass. 

shed, as a rule, nor are the outgrowths branched. Yet every rule has its exception, 
exemplified in the case of the Cavicornia by the prongbuck, which sheds its branched 
horny sheaths or antlers. The relationship of this "antelope" to other mammals is 
still, however, an unsolved problem. 

So numerous and so varied, at least in external form, are the animals called 
antelopes that "authorities" differ as to the definition of the family. Employed even 
in its widest sense, it includes few denizens of continents other than Asia and Africa; 
it is most abundantly represented in the latter area, although Asia has its hordes of 
gazelles, central Europe and central Asia their chamois, and America the Rocky 
Mountain goat. So far as is now known, antelopes did not reach Africa till Pliocene 
or Pleistocene times, but since their arrival they have thriven wonderfully. Travelers, 
keepers of zoological gardens, sportsmen, and naturalists have made us familiar with 
the brilliant colors, the polished, straight, ringed, or twisted horns, and the great 
variety and range in size of these creatures. Some are but little larger than rabbits; 
others have the bulk of oxen; some are slender, graceful, cursorial dwellers in 


deserts; others, sluggish, phlegmatic, semiaquatic occupants of swamps. The names 
bongo, kudu, oryx, saiga, steinbok, and hartebeest have so caught the imagination 
of makers of books and stories as to have become almost household words. Like 
most of the modern fauna of that country, the antelopes are not of African but of 
more northern derivation. The oldest yet known are gazelle-like creatures found 
in the Miocene of Europe and India. The eland, kudu, and hartebeest appear 
to have been in India during a part, at least, of the Pliocene. At this same time 
antelopes of some sort seem to have invaded North America, but the fossils so 
far found are too fragmentary to be understood. The Pleistocene saw their migration 
as far south as Brazil and the Argentinian pampas, but they failed to get a foothold on 
the southern continent. 

According to some zoologists, the chamois of Europe and Asia and the Rocky 
Mountain goat are intermediate between the antelopes and the goats. Others main- 
tain that our shy, surefooted mountain climber is no goat but a true antelope. What- 
ever its relationships, the mountain goat was probably a Pleistocene immigrant from 
Asia, for bones belonging to a member of its genus have been found in caves in Cali- 
fornia. True goats, on the other hand, were imported by man, never having risked 
the Bering passage themselves. 

Some students of mammals divide the hollow-horned ruminants into three fami- 
lies, the first containing the prongbuck and its extinct allies (Antilocapridae), the 
second the antelopes (Antilopidae), and third the cattle, including bison, sheep, goats, 
et cetera (Bovidae). This practice is followed here, although no attempt is made to 
enter upon the long, involved discussions which would be necessary if one were to 
offer definitions of the second and third families. Others, with considerable justifica- 
tion, place the antelopes with the bovids in the family Bovidae. There is some indica- 
tion that the Miocene antelopes may have been ancestral to all later Antilopidae and 
Bovidae, that the Pecora branched into the two characteristic groups in the Oligocene, 
and that the Pecora and Tylopoda (camels) had common ancestors in the Eocene. 
Lack of fossils prevents, at the present time, anything more than the general sugges- 
tion. True bovids appear to have been of Eurasian ancestry, none reaching North 
America till Pliocene or Pleistocene time. For that reason they will not be discussed 
here but will be referred to on a later page. 

Thi$ old world of ours has been a long time in the making. Only recently was 
the time ripe for the multiplication of the vegetarians. When the opportunity came, 
they seized it. But man in a few thousand years has nullified the effects of millions 
of years of evolutionary change, and his efficiency as a destructive factor has been 
accelerated in recent decades. Most of the vegetarians are doomed, despite the efforts 
of a few farsighted naturalists to save them. 


'Tis he, I ken the manner of his gait; 
He rises on the toe: that spirit of his 
In aspiration lifts him from the earth. 

Troilus and Cressida, Act IV, scene 5 

Many question the value of human pedigrees and deride as ancestor-worshipers 
those who are interested in genealogical pursuits; yet these same people will pay a 
premium for a pedigreed dog or horse or bull. All are agreed, however, that one of 
the most laudable pursuits of the paleontologists is the tracing of lines of descent. 
This is, in fact, the only possible test of evolution. If the idea is sound, it should be 
possible to find in the rocks the ancestors of many, if not all, of the living animals 
and plants. Realizing as he does the manifold imperfections of the geological record, 
the student of fossils hardly expects perfect results. Yet much has been accomplished. 

Invertebrate fossils are much more abundant than remains of vertebrates and 
have allowed the construction of numerous phylogenies. But these are not sufficiently 
spectacular to arouse popular interest. The description of the ancestors of familiar 
mammals has appealed to more people. As examples of this sort of work, it is most 
convenient to survey briefly the evolutionary history of the horse, the rhinoceros, 
and the camel. 

Darwin put together in a coherent theory the rather nebulous ideas of evolution 
which were a part of the thought of his day. The genealogy of the horse, as de- 
ciphered by O. C. Marsh, was hailed as the first actual proof of the theory. Although 
much has been done in tracing other lines, this family remains at the present time 
the best known, far better understood now than when Marsh first pointed out the 
salient facts of its evolution. The story has been told so often that it is familiar 
to all, but a summary is inevitable. The present account follows that of the late 
W. D. Matthew, America's most learned vertebrate paleontologist. 

The specializations of the skeleton of the modern horse are those of a large, 
swift-running, hoofed quadruped which depends for its food on the dry grasses of 
open plains. The skull and neck are long, as are the legs and feet; the gait is un- 
guligrade that is, the animal walks on the last or ungual phalanx of its toe. The 
dentition is somewhat reduced, canines being absent in females and the first premolars 
vestigial in both sexes. All the teeth are hypsodont, roots forming only with advancing 
age. The incisors are strong, powerful croppers, each with a deep cement-filled pit in 


the center. The premolars and molars are alike almost rootless prismatic teeth that 
continue to grow out as they are worn down, the complex infolding of the enamel into 
areas of dentine and cement producing a grinder which cannot be worn smooth. The 
eyes are far back, a position necessitated by the depth to which the teeth are inserted. 
The legs are long slender pendula, actuated by masses of muscle in fore quarters and 
flanks. Grooved articulations between the bones allow fore and aft and prevent lateral 
movement. The mid portions of ulna and fibula are lost, the vestigial ends being 
coossified with radius and -tibia respectively. Only the third digit of the foot is func- 
tional, the first and fifth toes having been entirely lost, whereas the second and fourth 
are represented by "splints," remnants of metapodials. The metapodial of the third 
toe is the long, large cannon bone. The ungual phalanx is broad and covered by a 
large horny hoof. 

More than half a century and hundreds of thousands of dollars have been spent 
in finding the links between the modern horses and their little four-toed ancestor, 
Eohippus (Fig. 140). Several species of this Lower Eocene genus are now known, 
animals varying in size from that of a cat to that of a small fox; that is, eighteen 
inches to two feet long. All had the complete dentition, the premolars being small 
cutting teeth, simpler than the molars. The latter had low crowns, each with two 
or three pairs of blunt, conical cusps, the whole entirely unlike the complicated pattern 
of the teeth of the modern horse. The gait of Eohippus was digitigrade; probably 
pads were present under the toes, as in the carnivores. The bones of the lower arms 
and legs were separate from one another. Four toes, all functional, were present on 
the front feet, whereas on the hind there were three useful and two vestigial ones, 
the latter represented only by tiny nodules. The orbits were at about mid-length of 
the skull. 

Eight genera show characteristics which illustrate the stages intermediate between 
Eohippus and the modern genus Equus. Many other kinds of horses are known, but 
these suffice to demonstrate the chain of significant changes. 

Orohippus of the Mid-Eocene shows the first change from the ancestor, for the 
vestigial toes have disappeared from the hind feet, the third toe is slightly larger on 
all four feet, and the fourth premolar is molar-like. The third stage is represented 
by the Upper Eocene Epihippus, which is of larger size and has two molar-like pre- 
molars. All of the Eocene horses are small; all have four functional toes on the 
front feet and three on the hind. Moreover, all walked on the fingers and toes, not 
on the tips of them; that is, they were digitigrade. 

Mesohippus (Fig. 132) of the Lower and Middle Oligocene shows a considerable 
advance, since the outer (fifth) toe of the front foot is reduced to a splint. The third 
digit of all feet is elongate and strong, being much the most important, although the 
two lateral ones still remain functional. Three premolars are molariform, and the ulna 
and fibula are attached to radius and tibia. The animals were about the size of 


sheep, up to about forty inches in length. Miohippus of the Upper Oligocene was 
somewhat larger than Mesohippus, with feet a little longer and stronger. Oligocene 
horses, then, had short-crowned teeth with a rather more complicated pattern than 
those of the Eocene, three-toed, digitigrade feet, and coossified lower limb bones. 

The Lower Miocene Parahippus (Fig. 133) is distinguished as the first among 
the horses to have cement on the crowns of the teeth, which show increase in height. 
The second and fourth toes were vestigial, the third elongate. Horses now began to 
walk on the ends of their middle toes. Merychippus of the late Miocene is repre- 

FIG. 131. At left, sketches showing the principal stages in the evolution of the hind foot of 
the horse. The tallest is E quits, then Merychippus, Miocene; next Mesohippus, Oligocene; and 
last, Eohippus, Eocene. At right, three stages in the evolution of the cheek teeth. Below, Eohippus\ 
in middle, Mcsohippus\ at top, Merychippus. From Matthew and Chubb, Evolution of the 
Horse. By permission of the American Museum of Natural History, New York City. 

sented by several species which in successive zones show higher and higher crowns on 
the teeth, the cement forming a heavy coating. Miocene horses have three toes, but 
the lateral pair are so much reduced that the animals walked on the tips of the 
middle toes only, as modern horses do. The teeth are high-crowned, the eye is back 
of the middle of the head, and increase in size has continued. The Upper Miocene 
and Lower Pliocene Pliohippus has cheek teeth three times as high as wide, and the 
side toes are actually reduced to long splints in some species. These animals were 
larger than Miocene horses. 

Plesippus of the Upper Pliocene forms the link between Pliohippus and the true 
Equus of Pleistocene and Recent times. Nearly as large as a modern horse, it differed 
from it only in small details of the structure of teeth and feet. Pliocene horses were 
practically one-toed animals, with most of the characteristics of modern ones, dif- 


fering chiefly in the longer splints of the lateral toes, the somewhat shorter teeth, and 
the smaller size. 

All of this evolution took place in North America, although the ancestral Eohip- 
pus may have been an immigrant from the Old World, for it seems to be identical 
with the Hyracotherium, long since described by Sir Richard Owen from bones 
found in the Eocene London clay. 

The horse is familiar to all, because he has been enslaved by man. Numerous 
pictures on the walls of caverns show that he was an object of human interest to the 
first known race of our species, Homo sapiens, some twenty thousand years ago, but 
that interest seems to have been evoked by the tastiness of his flesh rather than his 
quality as a beast of burden. 

Known to man equally long but apparently less closely associated with him, for 
it was only rarely depicted, was the rhinoceros. Never domesticated, perhaps rarely 
killed, the great woolly rhinoceros was probably looked upon with awe. His red 
ochre portrait on the wall of the cavern of Font-de-Gaume in Dordogne may com- 
memorate some tragedy in the history of the local tribe. Later single-horned rhi- 
noceroses are supposed to have inspired the myth of the unicorn. Existing members of 
this group occupy restricted areas in Africa and southern Asia. Those living in 
India and Java have only a single horn; in Africa and Sumatra are found similar 
creatures with two horns in tandem, one above the nose, the other on the forehead. 
The present distribution of these animals is a relatively modern one, for in Pleistocene 
and early post-Pleistocene times they were abundant in Europe. Still earlier they 
were even more widespread, for from the Oligocene onward they roamed through- 
out Europe, Asia, and North America. In the Eocene of the last are found the 
oldest of them. 

The modern rhinoceros bears little resemblance to the horse. It is a much less 
graceful creature, with relatively short head, neck, and legs, long heavy body, and 
thick, hairless hide. The feet have three toes each, the median one the largest. Most 
rhinoceroses dwell in marshes and forests, feeding on leaves and shrubs; that is, they 
are browsers. The dentition is reduced in all; the full series of cheek teeth is re- 
tained, but canines are absent and the number of incisors is less than that of primitive 
mammals. The cheek teeth have a characteristic pattern, that known as the lopho- 
dont, more easily illustrated than described but far simpler than that of the horse. 
Premolars and molars have the same crests. The horns, although long and con- 
spicuous, have no bony cores, their presence being indicated solely by the thickening 
and roughening of the nasal bones. 

The American rhinoceroses, represented by fossils from Eocene, Oligocene, Mio- 
cene, and Pliocene deposits, appear to fall readily into three major groups. First 
and oldest are the light-boned, cursorial forms, the hyracodonts of the Eocene and 
Oligocene. The oldest of these belong to the Mid-Eocene genus Hyrachyus 9 some 

FIG. 132 Mesohippus barbouri, a three-toed Oligocene horse, twenty-two 
inches high at the shoulders. Skeleton in the Museum of Comparative Zool- 
ogy, Harvard University. 

FIG. 133. Parahipptts wyomingensis, a Lower Miocene horse, thirty-eight 
inches high at the shoulders. Note reduction of lateral toes. Skeleton in 
Museum of Comparative Zoology, Harvard University. 

FIG. 134. Stenomylus, a slender, long-legged, swift-footed Miocene camel. 
Photograph by George Nelson of the specimen in the Museum of Comparative 
Zoology, Harvard University. It is twenty-six inches high at the shoulder. 


species of which are as small as a fox, others as large as a tapir. It is probable that all 
later rhinoceroses were descended from members of this group, although the genea- 
logical lines have not yet been proven in detail. Hyrachyus itself may have been de- 
scended from Homogalax, the reputed Lower Eocene ancestor of the tapirs. All 
species of Hyrachyus had long bodies, short necks, four toes on the front feet, and 
three bn the hind. The skull must have been hornless, for the nasals are thin. All 
of the primitive forty-four teeth were present, the premolars smaller than the molars. 
Only the upper true molars were lophodont, of the true rhinocerine type, the lower 
ones resembling those of tapirs. 

Somewhat larger cursorial forms are found in the Upper Eocene, animals that 
had one less toe on the front foot. These serve to connect with the Oligocene Hyraco- 
don, the last of this line. Hyracodon was somewhat taller and heavier than a sheep, 
with a large head and stocky limbs. It was, however, proportionally much lighter in 
build than any modern rhinoceros, with a longer neck and more slender bones. Each 
foot had three toes, the median larger than the others; all premolars except the first 
were molariform. In many respects Hyracodon resembled the contemporary horse, 
Mesohippus. Both had three toes, light bones; both were fleet of foot. But Hyracodon 
died out in the Oligocene, whereas Mesohippus left numerous descendants. Why? 
Apparently because one was a little more successful than the other in eating the most 
abundant food then available, the grasses of the plains. 

According to the latest views, a Lower Eocene mammal similar to Homogalax 
or Eohippus was probably the ancestor of the rhinoceroses, the tapirs, and the horses. 
Some of the descendants of Eohippus deserted the forests. In these the teeth became 
modified by the grinding of dried grasses, the limbs changed as a result of the mechani- 
cal reaction to running on hard ground, and the foot was lengthened. Others re- 
mained in the forests, and in them the teeth were modified by use in the constant 
chewing of leaves and twigs, the legs changed as the result of impact on an irregular, 
relatively soft soil, and the lower arm and leg bones, rather than the feet, were elon- 
gated. During the Oligocene the drier climate caused a great reduction of the area 
of forest and a corresponding increase of grassy plain. Running rhinoceros was forced 
to compete with running horse. Teeth, not legs, decided the issue. 

The oldest true rhinoceros now known is Trigonias, a Lower Oligocene animal 
with four toes on the front feet and three on the hind, the little finger small but 
functional. Except for the loss of the lower canines, it has the dental formula of the 
Eocene Hyrachyus. The limbs were more massive, but, as in the earlier form, there 
were no horns. These animals, known as the aceratheres (Fig. 135), persisted in 
North America till Mid-Pliocene times. 

Somewhat nearly allied to them were the first to bear horns, the diceratheres 
(Fig. 136), which seem to have had small ones on the anterior tips of the nasals. 
Since the horns are represented merely by rugosities on the bones, it is probable that, 


one remembers something of the geography of his school days, the llamas of Andean 
South America may come to mind; but the name has no association with North 
America. Nevertheless, this continent was the home of the camels throughout their 
history. It was the nursery of their infancy, the romping ground of their adoles- 
cence, and the Eden whence in the fullness of their strength they were at last expelled, 
"to labor and sorrow all the days of their lives." Eocene times saw their inception in 
North America; only recently, probably since the end of the glacial period, did they 
disappear. Not until the Pliocene, so far as present records tell, did any of them 
migrate to Asia or South America. 

Although they retain some primitive characteristics, the camels are among the 
most specialized of the artiodactyls. Their position in this respect is comparable to 
that of the horses among the perissodactyls. The larger camels, it is true, are rather 
awkward, lumbering animals, not to be compared in fleetness with the antelopes, 

FIG. 136. Palatal view of the skull of a Dicer athcrium, showing the typical 
lophodont teeth. One-ninth natural size. From O. A. Peterson. 

but some of the Mid-Tertiary representatives of the group must have rivaled the 
gazelles in cursorial powers. The two most obvious specializations of the modern 
camels are the elongation of the vertebrae of the neck and the growth of one or two 
"humps." The latter are reservoirs of fat, without any unusual skeletal support. 
Hence the presence or absence of a hump on any extinct camel cannot be determined. 
The general impression, however, is that this method of storage of potential energy 
is a modern development; it is doubtful if any ancient member of the group had a 
hump. The dentition is reduced to thirty-four by the loss of the first and second 
upper incisors, one upper and two lower premolars. The outer upper incisors, the 
canine, and the first premolar are all short, sharp teeth, widely spaced in the elongated 
muzzle, and also separated from the selenodont cheek teeth by a wide diastema. The 
molars are high-crowned but not hypsodont. Ulna and fibula are vestigial, the 
former fused to the radius, the latter represented only by its lower end, which forms 
the so-called "malleolar" bone. The foot is greatly elongated. There are only two 
toes, the third and fourth remaining after the loss of the outer ones. They are of 
equal size, somewhat divergent, supported by pads in an unguligrade position. The 
corresponding metapodials are long, fused into a cannon bone which has a A-shaped 


as in the living forms, they were composed of densely matted hair rather than of bone 
or horn. They were placed side by side, not in tandem as in the modern double- 
horned animal. This group had a brief existence, from Mid-Oligocene to Lower Mio- 
cene. Probably most readers have seen the remarkable slabs in various museums 
showing the bones of these creatures. All of them, containing thousands of limb 
bones, ribs, vertebrae, and skulls, have been excavated from a single layer in a hill 
on the ranch of Mr. Harold Cook at Agate Springs, Nebraska. 

These were the only double-horned rhinoceroses. The type with the horns on the 
median line appears to have been chiefly Eurasian. Teleoceras (Fig. 135), with a single 
horn on the nasals, was in America during the late Miocene and Pliocene, but was a 

FIG. 135. Above, the short-legged Teleoceras^ an Old World type of 
rhinoceros which reached America in Miocene times. Below, a primitive 
Oligocene acerathere. Redrawn after H. F. Osborn. 

short-legged, paludine type which could not have given rise to modern forms. It was 
also in Europe in Miocene times and may have been of Eurasian ancestry. 

Although primarily North American, some rhinocerine group reached the Old 
World in the Oligocene or earlier. Evolution appears to have proceeded there along 
several lines, although the records are as yet too scanty to allow connections to be 
made with either the modern descendants or the American ancestors. Asiatic locali- 
ties have furnished remains of the largest and most extraordinary of all, the gigantic 
Baluchitherium . (Fig. 137) of Baluchistan, Turkestan, and Mongolia. Seventeen 
feet high at the shoulders, with huge hornless skulls more than five feet long, the 
baluchitheres must have equaled the giraffes in ability to browse high in the trees. 
They had but a brief history, their remains being known only from late Oligocene 
or early Miocene deposits. 

When camels are mentioned, one's thoughts turn naturally to distant places; 
visions arise of the shifting sands and elusive oases of the Sahara or the romantic, 
almost legendary, caravans which conveyed the luxuries of the East into Europe. If 


notch at the lower end, separating the surfaces with which the proximal segments of 

the toes articulate. 

The oldest "camel" is Protylopus (Fig. 138, at left), an animal not larger than 
a jack rabbit. But how large a jack rabbit may be depends upon the veracity of one's 
informant; some which have jumped from their hiding places at the approach of 
the writer have seemed as large as deer. At any rate, the jacks are larger than the 
cottontail, perhaps as large as a "small dog." The comparisons made by vertebrate 
paleontologists in their efforts to enable the student to visualize the extinct animals 

FIG. 137. Restoration of Baluchitherium\ height, 17 feet 9 inches at the 
shoulder. Redrawn after Charles R. Knight, American Museum of Natural 
History, New York City. 

are amusing in their indefiniteness. The largest titanotheres were as large as "small 
elephants"; the earliest horses were as tall as "large cats"; the oreodonts were as big 
as "tall pigs." Perhaps, after all, these statements are the best that are possible, for 
different species of extinct animals varied in size just as individuals of modern species 
do. The Upper Eocene Protylopus may or may not be the direct ancestor of modern 
camels, but it shows a condition which is primitive in the group. All forty-four teeth 
are present, in a continuous series. The molars are short-crowned (brachydont) and 
square, not elongate. The pattern, however, is selenodont. Ulna and fibula are com- 
plete, not coossified with their companion bones. The hand has four fingers, the 
outer ones smaller than the inner, and on the feet are dewclaws, small, vestigial 
remains of the second and fifth digits. 


Even more primitive than Protylopus is Diacodexis of the Lower Eocene, con- 
sidered by Matthew to be the most primitive artiodactyl known. It is not, unfortu- 
nately, of such structure as to indicate that it was ancestral to all even-toed mammals, 
but it may represent the stem toward which all selenodonts, namely, oreodonts, camels, 
deer, and ruminants, converge as their histories are traced further and further back- 
ward: Diacodexis was not so large as a jack rabbit, but more nearly comparable in 
stature to "bunny," the semidomesticated rabbit. Not only had it the primitive 
forty-four teeth, but the cheek series were of the bunodont rather than the selenodont 
pattern, the molars actually tritubercular. Both front and hind feet retained four 
toes, although the outer ones of the latter were slender, the only reason for excluding 
this animal from the proud position of ancestor of all artiodactyls. Its immediate fore- 

FIG. 138. At left, Protylopus, the possible Eocene ancestor of the camels. 
One-third natural size. At right, the Oligocene Poebrothcrium. Two-fifths 
natural size. Both from J. L. Wortman. 

bear may have had fully developed second and fifth digits on the hind foot, in which 
case it would be the connecting link between artiodactyls and creodonts, for Diacodexis 
retains many of the characteristics of the latter group. 

The Oligocene Poebrothcrium (Fig. 138, at right) is truly ancestral to existing 
camels, whatever may be the ultimate interpretation of the Eocene animals just 
mentioned. About as large as a sheep, it had a complete dentition, although not in a 
continuous series, for there are spaces between the anterior teeth. Incisors and canines 
are all of approximately equal size, the lower incisors erect and sharp, not broad and 
procumbent as in the modern animals. The molars are selenodont but short-crowned. 
Skeletons have been recovered from several different levels in Oligocene strata, each 
higher bed furnishing somewhat taller and larger individuals, but all belong to the 
same genus. According to W. B. Scott, all were slender, "with small pointed head, 
long neck and body, and long, very slender limbs and feet ... the forearm bones 
were fully coossified, and in the lower leg only the two ends of the fibula remained." 
Although the lateral toes were reduced to vestiges, only two remaining functional, 
each with a deerlike hoof, the metapodials were not fused into a cannon bone. 

The Oligocene was the period of childhood for this group, but the Miocene was 


that of their lusty youth, the time during which they reached man's estate. According 
to Matthew, the Lower Miocene Oxydactylus, the Mid-Miocene Protolabis, and the 
Upper Miocene Procamclus are the genera which carry the blood in direct line toward 
the modern Camelus. Oxydactylus did not differ greatly from Poebrotherium except 
that it was taller, with longer legs, neck, and skull. The molars had high crowns 
but were not hypsodont, and the feet had deerlike hoofs, without a pad. Its successor, 
Protolabis, still retained the whole set of teeth, and a cannon bone had not yet been 
formed. Procamelus (Fig. 139), however, was as large as a modern llama, had cannon 
bones, and had lost the first two pairs of upper incisors, although it retained all the 
premolars. It may have been ancestral to both the true camel and the llama. Not 
till Upper Miocene times did the molars become hypsodont and the lower incisors 

FIG. 139. A Miocene camel, Procamelus, with dentition much like that of 
Recent ones. One-twelfth natural size. From Earl Douglass. 

The Pliocene camels were racially fully adult, most of them as large as any mod- 
ern camel, and at least one was larger. The end of the Miocene apparently marked 
the parting of the ways of two lines, so far as size is concerned. The llamas, migrating 
to South America, where they now occupy the uplands, failed, perhaps because of the 
rigorous environment, to reach greater size than Procamelus. The toes remained 
separate, each with its own pad. In contrast to this conservatism, loss of teeth 
proceeded further than in Camelus, two upper incisors, two upper, and two lower 
premolars being absent from each jaw of the modern llama. The large Pleistocene 
Camelops has the same formula, being in this respect somewhat more specialized 
than the camel which has survived. Pliauchenia is a genus which includes many 
Pliocene forms. Some of the species are relatively small, and may connect Procamelus 
and the llamas; others are large, and consequently in the line leading to modern 
Asiatic types. Among these large creatures are some which, apparently for the first 
time in the history of the group, became adapted to life in deserts. The broad, poorly 
developed ungual phalanges indicate that, like the modern camels, the large Pliau- 
cheniae had large single pads on each foot. 

The history of the camel is in the main parallel to that of the horse. The Eocene 
ancestors of both were small, and there has been a constant increase in size, culminat- 
ing in the Pliocene. Concomitantly with increase in size there has been an increase 
in length of limb, accomplished principally by the elongation of the foot. Both have 


suffered loss of toes but have gained thereby in fleetness. In both the skull has become 
elongate, with loss of teeth between the incisors and molars. The latter are in succes- 
sive stages higher and higher crowned till at last they become truly hypsodont, 
almost rootless. In fact, convergence in the evolution of these two unrelated groups 
extends further than mere similarity of general trend. A horse and a camel from a 
formation of any particular stage of the Cenozoic are of about the same size, about 
the same length of leg, and, in all respects, of about the same degree of specialization. 
This suggests that evolution was controlled by environment, for in both lines the 
ancestors were browsers, living on leaves and shoots found in the forests, and the 
descendants grazers, more and more adapted to exist under the increasingly semiarid 
climates of the plains on which they lived. 

Collateral lines have purposely been avoided in the phylogeny of the camels, but 
this should not lead to the impression that camel evolution was along one line only 
or that camels as a whole reached the "top of their form" only when they achieved 
their greatest size in the Pliocene. As a matter of fact the survivors of the group 
appear to belong to a strain which was rather backward, conservative in its changes. 
True grazers did not appear in this line till the Upper Miocene, whereas the Lower 
Miocene Protomeryx seems to have been fully adapted for subsistence on grasses. The 
contemporary Stenomylus (Fig. 134) the "gazelle camel," was better adapted for 
rapid locomotion than any other member of the group, living or extinct. But these 
are representatives of phyla which for one reason or another were unable to survive. 

FIG. 140. Restoration of Eohippus, redrawn after Charles R. Knight's 
painting in the American Museum of Natural History, New York City. 


O! call back yesterday, bid time return. 

King Richard the Second, Act III, scene 2 

Mammalian history reached its climax just as man was emerging from obscurity. 
He was confronted in both Eurasia and North America by an extraordinary abun- 
dance of beasts, many of them superior to him in strength, speed, and cunning. Some 
were active, others potential, enemies. But, by and large, abundance of mammals 
meant abundance of food, provided only that it could be captured. And so man 
became a hunter. 

The North American Pleistocene fauna seems at first view to have been a 
strange one, for it contained a mixture of creatures now belonging to this and other 
continents and provinces. Wild camels and horses are now Asiatic in their distribu- 
tion, elephants are Asiatic and African, lions are African, and tapirs and sloths are 
Central and South American. All these animals were common in North America 
during the Pleistocene. Some of them had evolved here and have disappeared re- 
cently; others were immigrants during the Pliocene or Pleistocene. 

The most conspicuous members of the fauna were the elephants and the masto- 
don. The latter was the more primitive, the last survivor of a stock which entered 
North America from Asia as early as the Miocene. One of the chief differences be- 
tween the two is in the teeth. The cheek tooth of the mastodon is much simpler than 
that of an elephant, for, although large, it has definite roots and a crown, on which 
are a few bunodont tubercles. Wear removes the superficial layer of enamel and 
produces a pattern showing four or five broad areas of dentine alternating with 
narrow bands of the harder substance. The elephant tooth, on the other hand, is 
almost rootless (hypsodont) and contains numerous cross-crests, the upper edges of 
a series of plates in which enamel, dentine, and cement (bone) are arranged in 
alternating sheets. 

The oldest representatives of the elephant tribe (Fig. 141) are Moeritherium of 
the Upper Eocene and Lower Oligocene of the Fayum in Egypt, and Paleomastodon 
and Phiomia of the Lower Oligocene of the same region. These animals were less 
than half the height of modern elephants and had many more teeth. Moeritherium 
had thirty-six, and the others twenty-six. The second upper incisors, which were to 
become the gigantic tusks of later elephants, were only slightly enlarged in the first, 
although conspicuous in Paleomastodon and Phiomia. The molars of all three were 
of normal bunodont type, foreshadowing those of the mastodons. 

FIG. 141. Diagram to indicate the radial evolution of mastodonts and ele- 
phants. Mastodonts: A, Moeritherium\ B, Phiomia\ C, Amebelodon-, D, 
Palaeomastodon\ E, Miomastodon; F, Mastodon americanus; G, Stegodon. 
Elephants: H, Mammonteus (= Mammuthus = woolly mammoth); I, Parelc- 
phas\ J, Archidiskpdon\ K, Elephas. Simplified after H. F. Osborn. 

FIG. 142. The skeleton oE a Pleistocene elephant (Parelephas columbt) in 
the Amherst College Museum. Photograph by courtesy of W. E. Corbin. 

FIG. I42A. An imaginary Siberian Mammoth hunt, the animal being driven 
over a cliff. From a painting by Ernest Griset in the United States National 
Museum, reproduced by permission of the Secretary of the Smithsonian In- 


The Miocene seems to have been the time of greatest mastodont prosperity. Then 
the group suddenly spread into Europe, Asia, and North America, occupying vast 
areas never before trodden by elephantine feet. They flourished exceedingly, 
increased greatly in stature, and lost the useless teeth between tusks and molars. 
Most remained in the forests and adhered to the ancestral browsing method of 
feeding. Not till the Pliocene do true elephants, with grinders of the modern type, 
appear. Just when or where or how they became adapted to life on the plains 
and to a diet of grasses is still a mystery. The transition probably took place in 
Asia, but some of the largest and most specialized roamed North America in 
Pleistocene times. 

Least conspicuous of this group was the Siberian or woolly mammoth (Mam- 
monteus primtgenius) , well known in central and western Europe, where it was 
trapped and pictured by primitive man. It spread across northern North America, 
occasionally venturing as far south as North Carolina. That its normal habitat was 
in the cold regions of the north is shown by the dense wool beneath the coarse hair 
of the frozen carcasses found in Siberia, its native region. It was only nine feet 
high at the shoulder, but the related Columbian elephant reached eleven feet, about 
the size of the modern African Loxodonta. The Columbian elephant had a more 
southern range than the true mammoth. The best specimens have been found in the 
southern and western states, but it seems to have been at home all over what is now 
the United States and in parts of Mexico. An impressive skeleton is on exhibition 
in the museum at Amherst College (Fig. 142). Both this and the woolly mammoth 
have numerous (2730) thin lamellae in their molar teeth. Largest of all the North 
American forms was the imperial mammoth of California and other southern states. 
It was thirteen and a half feet high at the shoulders, rivaling the contemporary 
straight-tusked elephant (Palaeoloxodon antiquus) of southern Europe and northern 

The Pleistocene American mastodon was about as large as the woolly mammoth. 
It is, perhaps, the most widely distributed of all American vertebrate fossils, its re- 
mains having been found at numerous localities from Alaska to Mexico. The legs 
are comparatively short, and the head lower and more flattened than that of the 
true elephants. A coat of long coarse hair fitted it, like the Siberian mammoth, for 
life in cold regions, but it chose wooded areas rather than open plains. It survived 
the last of the glacial stages and seems to have lived until recent times. 

The mastodonts had been established in North America since the Miocene, but 
the elephants were new arrivals. Other immigrants from Asia during the Pleistocene 
were the mountain sheep, the Rocky Mountain goat, the musk ox, and the bison. 
The last were the only ones really to flourish, although musk oxen were once much 
more numerous and diversified than at present and at times came as far south as 
Kentucky. Great herds of bison grazed on the western plains, just as did their one 


survivor, the "buffalo," within the memory of men still living. In earlier days there 
were numerous species, some including individuals large enough to support horns 
with a spread of from seven to ten feet. As was mentioned in a previous chapter, 
deer had reached North America as early as the Miocene, but they were not abundant 
till the Pleistocene. There is some evidence that there were new immigrations from 
Asia at that time. 

Horses and camels, typical American animals, throve during the glacial age, 
but after having survived the worst nature could furnish in the way of cold and 
climatic changes they were blotted out from this continent before the advent of 
white men. Horses ranged in great herds all the way from Alaska to Mexico. All 
belonged to the modern genus Equus, but neither Equus caballus, the domesticated 
horse, nor its immediate ancestor was among them. There were at least ten species, 
varying in size from the pygmy horse of Mexico to Equus giganteus of Texas, larger 
than any modern draft horse. There was a forest-dweller at the time, an animal of 
moderate size, but the favorite habitat was the grassy plains. 

Peccaries, the American representatives of the swine, had probably been in this 
country since the Miocene, but, like some other groups, they became most abundant 
in the Pleistocene. They still remain in the region from Texas to Brazil, though they 
have been driven from the more northern part of their range. 

Tapirs passed through most of their evolution on the northern continent but are 
now found only in Central and South America and southern Asia. Some of them 
came as far north as Pennsylvania during warm interglacial periods. Two species 
then roamed the United States, a large one now extinct and another smaller animal 
which seems to be the same as that still existing further south. 

Among the most striking of the immigrants are those from South America. 
Strangely enough, one is the Canadian porcupine, a rodent which has become so 
firmly associated with the north that knowledge of its origin comes to one as a sur- 
prise. Another rodent, the South American water hog, the largest existing member 
of its order, came north but failed to get a foothold. Even more typically South 
American were the great glyptodonts and ground sloths, now extinct but a con- 
spicuous part of the fauna of both continents during the Pleistocene. They mark 
the culmination of the peculiar order Edentata, represented nowadays by the hairy 
anteaters, tree sloths, and armadillos of subtropical and tropical regions of South 
and Central America. The armadillos, like the peccaries, are numerous as far north 
as Texas. 

The term "edentate" really applies only to the anteaters, for they alone are tooth- 
less. But tooth trouble is characteristic of all members of the order, even to the most 
distant ancestors. In general, incisors and canines are absent, and the cheek teeth 
are rootless pegs with no enamel on their surfaces. The number of molars is reduced 
in some, but, curiously enough, a few have more teeth than the normal placental. 


The case seems to be parallel to that of the whales, in which supernumerary teeth 
were added when the molars became simple and rootless. 

The edentates may have originated in North America, for the remains of primi- 
tive members of the group have been found in Paleocene and Eocene formations 
there, but their real evolution and differentiation took place in the southern conti- 
nent, where they were early arrivals. An origin in North America, however, is not 
supported by the present opinion of most paleontologists. Following Matthew's idea 
that the center of placental evolution was somewhere in the Arctic region and that 
the paths of migration radiate in all directions, it has been suggested that the early 
Tertiary specimens just mentioned were stragglers left behind as the group pushed 

FIG. 143. South American types of edentates which were on die northern 
continent in Pleistocene times. A Glyptodon in foreground, Nothrotherittm at 
right, and Mylodon in middle. The ground sloths are redrawn after restora- 
tions by Charles R. Knight under the direction of Chester Stock; the glypto- 
dont after R. Bruce Horsfall. 

southward toward its ultimate home. In this particular case, however, Matthew's 
idea of radiation does not apply, for the edentates never reached Eurasia or Africa. 
The Old World pangolin and the aardvark are toothless but are not members of the 

Whatever may be the ultimate decision on this matter, all will admit that the 
oldest edentates now known had departed considerably from the primitive placental 
condition. Only two incisors were present, and the cheek teeth were peglike, with 
mere vestiges of enamel. The best-preserved specimens are Mid-Eocene in age. 
Much may yet be learned as Paleocene deposits are explored. Although the North 
American Eocene specimens retain no traces of bony scutes, their osteology suggests 
a relationship to the armadillos, and parts of the armor of these creatures have been 
found in the Eocene of Patagonia. 

Modern armadillos do their best to Drove that there are nnrnhle pYrenrinnc tn the 


statement that sluggishness begets armor. They have broad shields of bony plates 
over shoulder and rump, and the area between is protected by narrow transverse 
bands, the whole covered with horny scales. Although the animals are almost com- 
pletely covered by a flexible coat of mail, their short legs are capable of carrying them 
across the open plains on which they live with great rapidity and of scratching out 
burrows with equal facility. They are not "choosy" about their food so long as it 
is tender. Insects, worms, small animals, and "root crops" appeal to them; in short, 
they are omnivorous. They have been abundant in South America since Eocene 
times. They reached their maximum size and differentiation in the Pleistocene, cul- 
minating in a form as large as a rhinoceros. This swashbuckler invaded North 
America but, after a short and glorious life, disappeared. Among his numerous 
relatives still surviving in the old home there is, however, one a yard in length and 
with fifty pairs of teeth. Hardly an edentulous edentate. 

The glyptodonts were the more phlegmatic and more fully armored cousins of 
the armadillos. Their shells, a mosaic of small polygonal bones fitted edge to edge 
like the plates of an echinoderm, were solid, inflexible, turtlelike. Some had a bony 
casque protecting the head, and rings of plates on the tail. A few had spikes on the 
caudal appendage, reminiscent of those of the stegosaurs. The legs were short, the 
claws broad, the jaws deep, with numerous ineffective teeth. These creatures could 
hardly have been burrowers, but on the other hand they could not have been really 
active. Yet it seems that they found food in abundance, for after their first appear- 
ance on the southern continent in the Oligocene they became common in the Mio- 
cene and abundant in the Pliocene and Pleistocene. In the latter epoch they reached 
their maximum size, with a length of nine or ten feet, and their greatest geographic 
distribution. Their remains are not common in the United States, although some 
reached the southern states and one adventuresome individual visited Atlantic City. 
All that has been found of him is a heel bone. 

Less is known of the history of the hairy edentates than of their armored rela- 
tives. Neither the tree sloths nor the hairy anteaters can trace their ancestry. They 
are unfortunate in that they have lived in regions where records were not kept. The 
ground sloths, although now extinct, seem to have belonged to the nobility, for their 
pedigree extends back to the Oligocene. As with modern noble families, a bit of 
tradition seems to be mixed with facts, for it is not till the Miocene that the group 
emerges from obscurity. As in the armored edentates, maximum size, differentiation, 
and distribution were reached in the Pleistocene. The large creatures are the ones 
in which we are chiefly interested, for their remains are common in southern United 
States, and not rare as far north as the states of Washington, Idaho, Kentucky, and 
Pennsylvania. Thomas Jefferson fathered a fabulous story of a gigantic bear which 
had once roamed the mountains of the Allegheny plateau; by avocation an amateur 
paleontologist, he had acquired huge bones from West Virginia which formed the 


basis of his tale. They were scientifically described as Megalonyx by Richard Harlan, 
who showed that they really belonged to an un-bearlike animal similar to a hairy 
ground sloth which Cuvier had described from South America in one of his first 
paleontologic papers. 

Most of the specializations of the ground sloths are associated with their ability 
to assume a semierect posture. The hind legs are short and massive; the feet ex- 
tremely awkward, large, and flat, with only one clawed toe; the pelvis a huge basin 
capable of supporting a great visceral mass. The long arms show that the animal 
was really quadrupedal, but the wrist was so twisted that the weight must have been 
borrie on the knuckles of the outer fingers. All the digits are clawed, useful in drawing 
branches and leaves to the mouth. The ground sloths were vegetarians and browsers 
though their teeth were soft and incompetent. 

Despite the various absurdities in their construction, nature dealt kindly with 
these creatures until comparatively recent times. Their giant, Megatherium, eighteen 
feet long, lived in both South and North America during the Pleistocene. Some- 
what smaller relatives survived even later, for specimens of Mylodon have recently 
been found in New Mexico with the tendons still undecomposed. Animals of its 
sort made the famous "human" footprints on the rocks in the prison yard at Carson 
City, Nevada. Another form (Glossotherium) seems to have been confined by man 
in a cave in Patagonia. Some naturalists have inferred that this animal was domesti- 
cated and that its cows were kept for their milk; the numerous droppings indicate 
that the animals were inhabitants of the cave for a considerable period of time. 

All of the Pleistocene creatures so far noted are of inoffensive sorts, herbivorous 
or omnivorous. Preying upon them in North America were numerous carnivores, 
some now extinct. They included minks, weasels, martens, raccoons, badgers, bears, 
foxes, wolves, coyotes, and pumas, which are still with us. Bears and badgers were 
new arrivals during the late Pliocene or the Pleistocene, having been strictly Eurasian 
previously. The most prominent of the extinct types were the saber-toothed tigers, 
repatriated descendants of Oligocene ancestors; they reached their culmination in 
size just before their extinction. Remains of these great catlike creatures have been 
found at many localities in the United States, and they were in the Old World as 
well. True cats, closely allied to the African lion, roamed the southern and western 
states. They are now gone, but some of the existing pumas are of a size not to be 
despised. Among the dogs, the "dire wolf" receives the premier place. A little larger 
than a modern wolf, he seems to have won through his name a reputation somewhat 
out of proportion to his real importance. 

It has not been possible in these few pages to express the full diversity of the 
Pleistocene mammalian fauna. O. P. Hay states that there were over six hundred 
species of vertebrates on the continent at the time, and that 60 per cent of them are 
extinct. I trust the reader has received the impression that North America then had 


numerous creatures which were newly immigrated, and that during the glacial 
period or shortly thereafter it lost a vast population. Some of the animals that dis- 
appeared were new arrivals, others descendants of ancient stocks. These changes are 
not easily explained. 

Till the Pleistocene most mammalian races increased slowly but continuously in 
cranial capacity, in size, and in adaptation to life in particular environments. Then 
many groups disappeared or were relegated to a minor position. It would be interest- 
ing to know definitely what were the controlling factors in the blocking of these 
evolutionary trends. Is there a limit beyond which size cannot increase? Undoubtedly 
there is, for an animal of great bulk requires more food for its maintenance than a 
small one, and, although there may be no limit to the amount of food available, there 
are definite limitations upon the amount which can be ingested and digested during 
a day. May not the evolution of some mammals and reptiles have stopped because 
each, according to its own method of feeding, had reached the greatest size pos- 
sible? This in itself is not a cause for extermination, but it brings animals to a con- 
dition in which they are vulnerable. It seems likely, too, that the extinction of some 
of the great races must be ascribed to geographical changes. In the case of the rep- 
tiles the worldwide mountain building which followed the Cretaceous appears to 
have been responsible for the downfall of the ruling houses of terrestrial animals. 
In the same way the glacial periods of the Pleistocene may have been, directly or 
indirectly, responsible for the fall of the mammals. 

The influence of the Pleistocene glaciation appears to have been chiefly indirect, 
for the various advances of the ice could hardly have been so rapid as to cause the 
extermination of warm-blooded, hairy mammals, capable of withstanding considerable 
cold and certainly well adapted for migration. Since the glaciation was bipolar, its 
effect was to drive animals both southward and northward. The effects of the re- 
frigeration were felt far beyond the limits of the areas actually invaded by the ice, 
the result being a southward shifting of the climatic zones in the northern hemi- 
sphere and a northward movement in the southern. Throughout the Pleistocene, an 
epoch in which there were three or four advances and retreats of the ice that pro- 
duced alternately cold glacial and warm interglacial stages, animals were forced to 
follow the shift of the climatic zone to which they were adapted or, if not that, at 
least one in which it was possible for them to survive. Changing climates altered the 
vegetation tremendously and consequently affected the food of the herbivorous ani- 
mals to a marked degree. Since plants, although more sensitive to changes in climate, 
do not migrate as rapidly as animals do, each stage of advancing ice put a double 
stress upon the mammals. Their crowding toward the equator greatly increased the 
population of the tropical and subtropical belts, and forced a spirited competition for 
food. Regions already fully occupied were called upon to support an enormous num- 
ber of immigrants. Not only did the incomers have to adapt themselves to a diminish- 


ing food supply if they were to survive, but, since the plants themselves were changing, 
they had also to accept alterations in their diet. 

It was unfortunate for North American mammals that dry land becomes nar- 
rower and narrower through Mexico and Central America to the bridge at the 
Isthmus of Panama. Although that bridge probably was higher and wider in Pleisto- 
cene times than it is now, it was the bottle neck of migration southward. Still less 
favorable was it as a route for animals coming northward. Yet it was used to a 
considerable extent by creatures moving in each direction. During the height of 
the glacial stages the climate of Pennsylvania and Iowa must have been much like 
that of Greenland today habitable in spots, but barely so. Only the most hardy 
could live there. Of all North America, only Florida, the Gulf States, the Southwest, 
and Mexico were really hospitable regions. Animals accustomed to life in these 
areas had to move still further south. Hence it was that pressure from the north 
forced many sorts of animals across the bridge of Panama. 

Throughout the Tertiary, South America was, in effect, an island. Although it 
was connected at various times with the northern continent, there was comparatively 
little migration across the isthmus except during the Eocene and Pleistocene. The 
tropical jungle of Brazil seems to have been almost as effective a barrier as though 
its area had been submerged. As a result of this isolation, the evolution of mammals 
in South America was practically independent of that at the north. A few stocks 
furnished the nucleus for a large and diversified population. We have already seen 
that some members of these groups traveled northward during the Pleistocene. Let 
us now list the principal ones which invaded the southern continent. 

Immigrant hoofed animals included horses, which apparently throve, for remains 
are found not only of Equus but of genera unknown in the north. Mastodons of 
various species crossed the isthmus, but, curiously, no elephants. Cloven-hoofed crea- 
tures were represented by deer, antelopes, peccaries, and many species of the camel 
tribe. Among the invading rodents were rats, mice, and rabbits. The descendants of 
some of them are still in South America; others failed to meet the competition of 
the native fauna. True carnivores, until then unknown in South America, entered 
the field and quickly ousted the native marsupial "wolves." In the list are a few 
bears, which is somewhat strange, for they had only just reached North America from 
Asia. With them came the saber-toothed tigers, jaguars, and pumas. Among the 
dogs were various small, foxlike wolves and, curiously, a dog of the same genus 
(Cyon) as the modern dhole of India a freak of geographic distribution. Smaller 
carnivores were the weasels, raccoons, skunks, otters, and a few others, but not all 
the North American fauna. The effect of such a sudden influx of flesh-eaters upon 
the great native population of herbivorous mammals must have been similar to 
that following the introduction of a family of weasels into a hen yard. 

South America is, from its shape, much better fitted to withstand the rigors of 


polar glaciations than its northern sister. Its southern end is not only considerably 
removed from the south pole but it has no great area, as compared with the vast 
surface of tropical Brazil. Ice covered Patagonia more than once, perhaps thrice, 
during the Pleistocene, urging mammals northward. They need not have crossed 
the equatorial belt, so far as we can see; yet some of them did. It may be that a 
period of glaciation in South America occurred at the same time as a warm inter- 
glacial stage on the northern continent, thus inducing a northern migration. Or 
possibly the influx of North American carnivores into the southern continent may 
have led to a northward drift in population. At first sight it may seem that the 
carnivores had much the best of this situation. It mattered little to them if plants 
disappeared from particular areas; the concentration of herbivores in restricted areas 

FIG. 144. Hypothetical restoration of the shovel-tusked mastodont, Ame- 
belodon, from the Pliocene of Nebraska. The lower jaws were nearly six 
and a half feet long. After a sketch by Erwin Hinckley Barbour. 

made for good hunting. In the earlier centuries of each ice advance beasts of prey 
doubtless flourished and acted their part, an important one, in the destruction of 
various groups of vegetarians. As the cold increased, however, the forests and thickets 
decreased. Lacking coverts, the pursuers became the pursued, for despite the absence 
of piercing claws and trenchant teeth, most large herbivores are far from being de- 
fenseless, or even of a pacific temperament. Lacking shelter or place of concealment, 
few of the solitary sorts of carnivores can compete successfully with herds of hoofed 
creatures Hence many of the carnivores, particularly those of the cat tribe, became 
extinct during the Pleistocene. Pack against herd is, however, another matter. 
Wolves were more successful until they met that defenseless two-legged creature 
who used weapons to subdue brute force. 

In spite of widespread sentiment to the contrary, it cannot truthfully be said 
that man has "interfered with the processes of evolution." Many writers, still un- 
consciously under the domination of the theory of the special creation of man, 
lament the fact that he has interfered with nature. Such and such would have hap- 


pened if it had not been for the intervention of man. It should be realized that man 
is just as much a product of nature, and of evolutionary processes, as any other 
animal. Because through the use of his large brain he has been able to subjugate 
all other creatures, it does not follow that he is set apart from other animals or that 
they were created for his use; he is merely one of the factors in the evolutionary 
forces 1 which control the history of life. Man is unique among animals in that he 
is able to exercise considerable influence over his environment. Nevertheless, he is 
not, and never will be, all-powerful. His skill has made many previously hostile 
regions habitable; on the other hand, some of his practices have brought desolation, 
through flood and erosion, to formerly fertile regions. Man is still being molded by 
his environment mentally, just as in Cenozoic times he was molded by it physically. 


He that overcometh shall inherit all things. 

Revelation, xxi, 7 

A survey of the chemical and physical properties of inanimate things in their 
relations to living matter led Professor Lawrence Henderson to the conclusion that 
all matter is biocentric. Certain it is that, in his egoism, man considers himself the 
creature to whom all else is subservient, and, with or without reason, has brought 
himself to the position of regarding the earth as Homocentric. Man's chief curiosity 
is, therefore, in icgard to his own origin. 

To write an account of the origin of man at the present time is to write a 
detective story. Nature, in one of her playful moods, has scattered clues, but, pixie-like, 
she seems to have delighted in giving just enough information to keep up the sus- 
pense. The romance of the search has attracted the keenest of morphologists, anthro- 
pologists, and paleontologists. Amateurs and professionals have given freely their 
time, resources, and mental abilities. Many solutions have been reached: some of 
them have been proved incorrect; others are strongly supported; but there are prob- 
ably few students who think that the case is closed. The many discoveries which have 
been made in recent years encourage one to believe that the story may finally be 
completed. Interesting as they are, however, the new clues have not been of much 
help in tracing the ancestry of man. Remains of Tertiary anthropoids, particularly 
of Miocene and Pliocene age, are badly needed. The world has been ransacked for 
such material, with meager results, but since there are still large areas in Asia and 
Africa which have not been explored half so intensively as those of North America 
there is still abundant opportunity for new discoveries. 

The conception that man is supernaturally set apart from all the other animals 
finds no support in the study of comparative anatomy. In fact, Huxley long ago 
showed that every bone and every part of man, save three little muscles of hand 
and foot, are present also in the higher apes, so that on anatomical grounds it is 
obligatory to class man, apes, and their allies together in a single order, the Primates. 
This does not mean that man is descended from the apes, as we know them, but that 
he must have had the same ancestry. 

As with horses and camels, it is necessary to understand the skeletal character- 
istics of man if one is to realize his place in the animal kingdom and understand 
how it has been reached. In some respects the skeleton is primitive. There are five 


digits, and the number of phalanges is the same as those of the Eocene mammals. 
The teeth are short-crowned, with few and low cusps, comparatively little reduction 
in numbers, and little specialization; they remind one of late Eocene mammals in 
which quadrangular molars had superseded the primitive triangular ones. 

The specializations have to do chiefly with adaptations to an erect posture and 
with the large size of the brain. The bones of the hind limbs are straight, the pelvis 
basin-shaped, and the spinal column doubly sinuate, all characteristics which help 
to support the body in a vertical position. Few appear to realize that in the whole 
vertebrate group there is but a single species, Homo sapiens, with its weight so dis- 
tributed that the axis is truly vertical. What plants and sessile animals do naturally 
in response to the stimulus of light has been accomplished but once by vagrant 
animals. Locomotion by the aid of two or more pairs of appendages has held the 
body in a more or less horizontal position. Dinosaurs achieved bipedality with the 
aid of a long tail which balanced the anterior part of the body; short-tailed birds 
have reached somewhat the same pose, partly because the greater portion of their 
bulk is concentrated in a relatively short body, but chiefly because the femora extend 
forward in an approximately horizontal position, bringing the lower legs beneath the 
center of gravity. The foot in man is specialized, although plantigrade. (In fact, it 
is a question whether the plantigrade mode of locomotion is primitive. Mammals 
with feet of this type are rare, and on the whole rather awkward; women realize 
the latter fact and employ high heels to make themselves pseudodigitigrade, a rever- 
sion to the semidigitigrade condition of the early Tertiary mammals.) The heel 
bone is elongated; the tarsals and metatarsals are so modified as to produce a strong 
arch; and the axis of the foot is not on the median line but on the inner side, where 
it passes through the big toe, which is more fully developed than in any other mam- 
mal. The outer toes are in process of reduction. The arms are relatively short, but 
man shares with the higher apes a striking feature, the power of rotating the thumb 
so that it is opposable to the remaining digits. This, however, is a characteristic of 
primates in general, probably inherited from the arboreal Paleocene ancestors. In 
contrast with the other primates man shows a decided loss of hair. 

Man has the same number of teeth, thirty-two, as the Old World apes and 
monkeys (i, c 4-> P-T> m 4)> but all are relatively smaller than those of apes. The 

L 1 L o 

canines do not project beyond the others, there is no diastema, and the series forms 
a semicircle rather than the U-shaped curve of most mammals. That reduction in 
the number is still in progress is shown by the fact that the first premolars and second 
incisors are small or may be absent, as are the third molars (wisdom teeth) of some 
individuals. This reduction in size of the teeth is correlated with the shortening of 
the muzzle and the retreat of the chin. 

These are some of the more obvious specializations of man. The characteristics 
of the skull deserve special mention because they are among the more striking, are 


most intimately connected with man's greatest asset, his brain, and because more 
information is available about their evolution than about that of the other portions of 
the skeleton. Most conspicuous are the shortness and height, the straight face, high 
forehead, and large, forward-directed orbits. The brain case is large, bulging upward 
and backward, and the foramen magnum is beneath rather than at the back of the 
skull. The nasal bones are small and short, the malar arches small, and the eye 
sockets partitioned from the temporal openings. In apes a very striking feature is 
the projection of the tooth-bearing portion of the jaws to form a muzzle; in man, 
particularly in the more civilized and higher races, there is no such prognathism. 
The shortness of the jaws appears to be correlated with the large size of the brain. 
As the anterior part of the cerebrum, the seat of the intellect, increased in size, the 
necessary enlargement of the skull in itself tended to straighten the face (or the 
facial angle); and, possibly to maintain a balance, or as a result of the changed 
direction in the pull of the muscles, the jaws shifted backward. Not all these charac- 
teristics are exclusively human, however, for some of the monkeys and apes have 
large brains, and hence large skull caps (calvaria), short faces, and occipital con- 
dyles as far forward as those of man. 

The order Primates is subdivided into three groups; the Lemuroidea, the Tar- 
sioidea, and the Anthropoidea, the last including the apes and monkeys. In the 
last suborder there are several families. The Old World anthropoids (catarrhines) 
are sharply differentiated from those of the New in two ways; they have narrow 
instead of broad nasal septa, which results in the nostrils' being close together and 
pointing downward, and they have either nonprehensile tails or none at all. The 
Simiidae, the highest family of Old World apes, share with the Hominidae the 
tailless condition and the dental formula, but have longer fore limbs than hind and 
differ in many other respects. These are, however, matters of proportion, not of 
the presence or absence of fundamental structures. The New World monkeys 
(platyrrhines) have flat nostrils, long tails (in some cases prehensile), and three 
instead of two premolars on each side. They are less closely related to man than 
the Old World series and probably had an independent evolution after Eocene 

The lemurs (Fig. 145, at right), which today are confined to the tropical forests 
of Africa, Madagascar, and southern Asia, are the most primitive primates. Their 
ancestors are found in Paleocene and Eocene rocks of North America and Europe. 
These early Tertiary primates have many characteristics in common with the in- 
sectivores of their day, as shown by their small size, elongate skulls, numerous teeth 
most of them have forty and by other structures. Some of the Eocene lemuroids 
are represented by fairly complete material. Perhaps the best-known American form 
is Notharctus (Fig. 145, at left), from the Mid-Eocene. It was a small animal with a 
skull that was about three inches long, a narrow brain case, and large orbits. There 


was no partition between the eye sockets and the temporal openings, but a continuous 
bar of bone behind the eye indicates the initial stage in the production of such a 
structure. There are forty teeth, the lack of one incisor on either side being the only 
reduction. The canines are larger and longer than the adjacent teeth, the premolars 
simple^ and the molars of the crushing type, with low tubercles. The molars of the 
older species of the genus are somewhat triangular, those of the later ones quadrangu- 
lar. Such teeth suggest omnivorous rather than insectivorous feeding. The feet 
show that the great toe diverged from the others, an indication of arboreal habits. 
Adapts is a similar but slightly larger and more powerful-jawed animal from the 
French Eocene. It is probable that the modern lemurs are descendants of animals 
not unlike these. They have been variously modified in the course of time but retain 

FIG. 145. A, the Eocene lemur, Notharctus. One-half natural size. B, a 
modern lemur, for comparison. One-half natural size. Both redrawn after 
W. K. Gregory. 

the primitive form, the grasping type of foot, and the long muzzle, though they 
have lost a few teeth, reaching the same dental formula as the platyrrhines. 

As early as the Paleocene, however, short-skulled primates made their appear- 
ance. They seem to have sprung from the primitive lemuroids and are in many 
respects so like them that only recently have paleontologists and osteologists recog- 
nized them as a special group, the tarsioids. This suborder gets its name from ex- 
tremely odd animals found on various peninsulas and islands from the Malay penin- 
sula to the Philippines. There are four species of the genus Tarsius, all queer little 
creatures with short, wide heads, big eyes which face almost directly forward, and 
highly specialized hind legs. 

Eocene tarsioids have been found in regions as widely separated as Wyoming, 
France, and Switzerland. They are somewhat longer-jawed than the modern animal 
but exhibit the same important characteristics, that is, a V-shaped arrangement of the 
dental rami and the rodentlike nature of the anterior teeth. Although in general 
form intermediate between lemurs and monkeys, all known tarsioids are too highly 
specialized to form connecting links. Tetonius (Fig. 146), the best-known American 
member of the group, is alleged to have been the most "brainy" creature of the 


Eocene. The skull is short and wide, rather catlike, but with enormous orbital open- 
ings. Its dentition, and that of its relatives, is reduced, in some species at least, to the 
same formula as in the South American monkeys. Unfortunately Tetonius appears 
to have lost all its lower incisors, the canines taking their place to form rodentlike 
anterior teeth. The European Eocene tarsioids are somewhat more monkeylike than 
the American ones, but none has yet been found which could have been ancestral to 
any of the later anthropoids, Nevertheless, there can be little question but that short- 
ening of the skull and increase in size of the orbits, with concomitant shifting toward 
the front, took place during Eocene times. Some as yet undiscovered tarsioid must 
have given rise to the European catarrhine monkeys and their more primitive South 
American platyrrhine cousins. As the brain increased in size and the sense of hearing 
became more acute, the breadth of the skull increased, tending to push the orbits 
into a more frontal position. At the same time the olfactory lobes of the brain, less 
important as sight and hearing improved, atrophied through lack of use, and the 
nasal region became smaller. This in turn brought the inner margins of the eyes 
closer together and shortened the face. The small size and unprotected nature of our 
distant ancestors compelled them to be timid folk, hiding during the day, abroad 
chiefly at night. Perhaps because they were arboreal rather than humble terrestrial 
quadrupeds, sight was more important to them than ability to distinguish scents. 

The belief that the tarsioids were ancestors of the anthropoids is strengthened by 
the study of two little lower jaws, the most important of the few specimens of primates 
which the reluctant Oligocene strata have as yet yielded. They were found in the 
Fayum desert southwest of Cairo and have been hailed as the jaws of the ancestors 
of monkeys, apes, and man. They show the typical tarsioid pattern in the V-shaped 
joining of the rami of the jaws, but the incisors are erect, not procumbent, and the 
formula is that of the catarrhines and man, i -, c , p -, m . One of these, Proplio- 
pithecits (Fig. 148 A), is so like the Upper Miocene and Pliocene Pliopithecus, and 
that in turn so like the modern gibbons, that there can be little doubt of its being 
ancestral to one or more lines of monkeys. The jaw is short, the canines are some- 
what larger than the adjacent teeth, the first premolar is bicuspid, and the last molar 
has five tubercles. The fact that it is a shorter jaw than that of the gibbons, with 
less prominent canines and less sectorial premolars, shows that the face has become 
more elongate, more prognathous, during the time since the Oligocene. On another 
line Propliopithecus may have been ancestral to the Miocene Dryopithecus and its 
allies, possibly to the African chimpanzee and gorilla, and possibly also to man. The 
single little imperfect jaw carries a heavy responsibility. The other Oligocene jaw 
is better preserved than that of Propliopithecus, for it retains both rami and is almost 
complete. It would be an excellent connecting link between the Tarsioidea and the 
Anthropoidea if there were only something with which it could be connected. One 
eminent student of the primates speaks of it indifferently as a tarsioid and a catar- 


rhine. The teeth are much like those of monkeys and apes, except for the fact that 
the canine is not enlarged, and there is no suggestion of a diastema. Yet Parapithecus 
(Fig. 147 A) has been, and still is, considered by some as a connecting link between 
the Eocene lemuroids and tarsioids with large canines, and later monkeys and apes 
with even larger ones. 

During Miocene times platyrrhine monkeys were present in South America, and 
catarrhine monkeys and tailless apes in Europe and southern Asia. Early in the 
Pleistocene, or before it, the higher anthropoids disappeared from Europe, although 
the monkeys remained in restricted areas, as they have to the present day at Gibraltar. 

It is not till the Pleistocene that man definitely comes on the stage. The first 
discovery to attract attention was that of Pithecanthropus erectus, found by Dr. 

FIG. 146. At left, skull of Tetonius, the most advanced Eocene tarsioid, 
after W. D. Matthew. At right, a restoration of the head of the same, by 
Edward R. Schmitz. 

Eugene Dubois near Trinil in Java in 1891-92. The material consists of the top of 
the skull, a femur, and three teeth. These bones are probably not those of a single 
individual, or even of one sort of primate. The femur belonged to a person having 
an erect posture and of average present-day size, but the skull has been shown by 
various tests to indicate a creature intermediate in cranial form and capacity between 
the higher apes and man. If the bones belonged together, this would indicate that 
in the evolution of man the erect posture and full height were attained before any 
great development of the skull, which is contrary to the teachings of embryology. 
The specimens were found associated with species of Hippopotamus, Elephas, and 
Stegodon, and are now believed to be of Mid-Pleistocene age. 

The cranial capacity of any skull in cubic centimeters is readily obtained by 
the simple process of filling the brain cavity with shot and then measuring the 
amount required. The average for modern Europeans is about 1450 cubic centimeters, 
though in exceptional cases it is as high as 2000 and as low as 1200, seldom less than 
the latter figure in normal healthy individuals. The gorilla has the largest brain 


among the Simiidae, and although there is a good deal of variation the average is 
about 500 cc., or one-third that of the normal European. The maximum is 650 cc. 
The cranial capacity of the skull cap from Java is estimated at 940 cc., intermediate 
between that of the gorilla and the modern European. Some among the primitive 
bushmen of Australia, however, have been shown to have as low a rating as 900 cc.; 
Pithecanthropus is definitely on the human side of the line. Lately it has been sug- 
gested that "he" was a lady, and that a male of this species may have reached noo cc. 
The calvarium is the most important of the various fragments of Pithecanthropus, 
for it proves that this hominid had a small brain, a markedly retreating forehead, and 
prominent ridges over the eyes. 

Sinanthropus petyngensis (Fig. 148 E) is the latest great addition to the human 
family. Filled fissures in Ordovician limestones at Chou Kou Tien, thirty-seven 
miles southwest of Peking, have yielded to the zealous labors of several geologists 
connected with the Geological Survey of China a series of teeth, well-preserved 
calvaria, and other fragments of men similar to* "Pithecanthropus. The brow ridges 
are equally strong, the vault is only a trifle higher, and the cranial capacity but little 
more, being about 1000 cc. Perhaps the chief interest in the find is that it establishes 
the Trinil race on a firm basis. Since numerous fossil mammals are present in the 
fissures, the date can be definitely fixed as early Pleistocene. Another important feature 
is that the caves contain numerous primitive flint implements and evidences of the 
use of fire. Sinanthropus is the oldest man yet found whose culture is known to have 
reached this relatively high stage. 

Another discovery was made in 1907 in a gravel pit at Mauer, near Heidelberg, 
Germany, under a thickness of about seventy feet of sand and loess. The associated 
animals include the straight-tusked elephant, a rhinoceros, and a lion, all African 
animals, but with them were remains of bear, bison, horse, boar, and ox. This mix- 
ture of African and northern species indicates a warm climate and, according to the 
best evidence, suggests the age to be the second or third interglacial period. The 
fossil is a lower jaw, with all the teeth preserved (Fig. 147 C). The most striking 
features are the massive ascending ramus, with a shallow notch at the top, and a 
retreating chin, both simian characteristics. The teeth, however, tell a different 
story. They are too small for those of an ape with a jaw of such size, there is no 
diastema, and the canines do not project beyond the other teeth. The molars likewise 
show patterns characteristic of primitive human races. The fossil was first named 
Homo heidelbergensis, but it has since been made the type of another genus, 
Paleanthropus. There are no processes for the attachment of the lingual muscles, 
an indication of poor development of speech. 

Next in order of age comes the Sussex man, but it is more convenient to pass 
on to the Neanderthal type, which is closely allied to the one just mentioned. Nean- 
derthal man (Fig. 148 F) has been known for a long time, the first skull having 


been found in 1856 in a cavern on the side of the Neander valley near Diisseldorf, 
Germany. More recently, numerous other remains of this type have been discovered, 
some of them associated with other fossils. Two skeletons found at Spy, near Namur, 
Belgium, were in strata containing remains of the hairy mammoth, and such evidence, 
together with the fact that most of the human relics are found in caves and rock 
shelters, suggests that the race flourished during the cold period of the last glacial 

The abundant material permits an accurate reconstruction of Homo ncander- 
thalensis, as he is technically called. Averaging only five feet (range 4'8"-5'3") 

FIG. 147. Jaws of (A) Parapithecus y Lower Oligocene; (B) Dryopithecus y 
Upper Miocene; and (C) Homo heidclbcrgensis^ Pleistocene; to illustrate the 
change from the A-shaped to the parallel-sided and thence to the fl -shaped 
arrangement of the teeth. A, after W. K. Gregory, modified from M. von 
Schlosser; B, after W. Branco; C, after O. Schoetensack. 

in height, he was not so tall as modern man. His posture was not absolutely erect, 
for the knee joint did not entirely straighten, and his enormous head was thrust 
forward. The trunk was short and thick; the arms were short, the hands large; the 
femur was curved, the lower leg short, and the foot clumsy. The skull is that of a 
brutish man, with heavy ridges over the eyes, a retreating forehead and chin, and 
large ocular and nasal cavities. The shape of the brain case indicates a high develop- 
ment of the posterior part of the cerebrum but less of that anterior portion which is 
supposed to be the region connected with the power of thought. In actual capacity 
the cranial cavity was larger than that of the average European, some skulls measuring 
1600 cc. 

This type of man appears to have existed for some thousands of years in Europe 


and other lands bordering the Mediterranean. Associated with his remains are found 
flint implements of the type known as the Mousterian, which indicate a stage of 
development that could have been reached only after long practice in the art of 
working flint. The study of implements has been actively pursued by archaeologists, 
particularly in France, Belgium, and England. As a result there has been built up 
a history of man based upon successive "cultures," as indicated by the flints found 
in those countries. Three broad classes are recognized: eoliths, either unchipped or 
with only casual working; paleoliths, flints chipped into shape but not polished; 
and neoliths, tools not only carefully chipped but polished as well. Those of Nean- 
derthal man are of the second class. They are finished on one side only, the other 
showing the surface resulting from spalling off the flakes. Nevertheless, they are 
adapted for various uses and prove that the Neanderthals were accomplished in the 
chase and that they could create the necessary weapons: they are credited with making 
the first hafted implements. There are also other evidences that they had achieved 
considerable civilization. Many of the skeletons were given burial by their fellows, 
and one found at Le Moustier in France had the skull resting on a flint plate of 
careful workmanship. The interment of weapons with the bodies shows a certain 
respect for the dead and indicates a belief in some sort of after life, if not (as Lull 
has inferred) actual immortality. 

Unlike all other fossil men is Eoanthropus, known from a fragmentary skull 
and the right half of a lower jaw (Fig. 148 C), with two teeth, the first and second 
molars, in place. The specimens were obtained by Mr. William Dawson from a 
small opening by the roadside at Piltdown, Sussex, England, and described by Sir 
Arthur Smith Woodward. It is difficult to determine their age, for fragments of 
mammals characteristic of the Pliocene and Pleistocene are mingled in the river-borne 
gravel. If Contemporaneous with the most modern of them, Piltdown man was 
probably not more recent than the third interglacial stage, since Hippopotamus and 
other subtropical animals occur with it. 

The skull is so fragmentary that those who have studied it have been unable 
to agree as to the proper reconstruction: estimates of its cranial capacity have varied 
from 1079 cc. to 1500 cc., and an intermediate figure of about 1300 cc. has finally been 
reached. It is not at all of the Neanderthal type, but has a high forehead like that 
of modern man. Aside from the fact that the bones are exceedingly thick, it is not 
peculiar. The jaw, however, is admitted by all to be more like that of a chimpanzee 
than like that of any man, living or extinct. This was recognized in the original 
description. The two teeth are like human molars, but the remainder of the jaw 
affords too much space to be filled by ordinary teeth. Hence, in his restoration of 
the anterior part, Smith Woodward made the canines large, like those of a chim- 
panzee, and allowed for a small diastema. The correctness of his view was demon- 
strated in a striking way the year after publication, when Dawson and Father Teilhard 


de Chardin, who were resifting the gravel at the spot where the jaw was found, found 
a large canine. It is twice as large as that of a man and almost exactly like that of a 
modern chimpanzee. This association seemed to many to be an unnatural one, so 
the jaw was attributed by some to a species of chimpanzee. The later finding of a 
few more fragments at a near-by site seems, however, to have convinced most of those 

FIG. 148. Sketches illustrating two of the possible lines of descent from 
Propliopithecus to Homo sapiens. A, Propliopithecus 9 after W. K. Gregory. 
B, Dryopithecus, after W. Branco. C, Eoanthropus, after Sir Arthur Smith 
Woodward. D, Homo sapiens, after W. K. Gregory. E, skull of Sinanthropus, 
lacking face, after A. S. Romer. F, Neanderthal man, after M. Boule. G, Homo 
sapiens, Cro-Magnon race, after R. Verneau. 

interested that skull and jaw belong together. Eoanthropus dawsoni> then, is to some 
people the missing link between man and the apes. The forehead is high, the brow 
ridge insignificant, and the brain large, all features of man, but the chinless jaw has 
the big canines of an ape. 

Toward the end of the last epoch of glaciation, modern man, Homo sapiens 
(Fig. 148 G), came into Europe. Whence he came is not known. The general 
opinion is that his was one of the waves of immigrants which have followed one 
another from undiscovered centers in western Asia; a few students look upon Africa 


as the possible homeland. The species had already differentiated into more than one 
race; just how many is still a subject of discussion. It is not likely that all came 
together, for they occupied separated regions. Whether they came while the Nean- 
derthals were still in possession and drove them out or whether they followed a line 
of least resistance and took possession of an area previously depopulated by "war, 
pestilence, and famine" is unknown. 

All early representatives of our own species differ from the Neanderthals in 
having higher foreheads, projecting chins, and a more erect posture. The best-known 
are the tall, or true Cro-Magnons, whose chief habitations were in northern Spain 
and southern and central France. In stature they may have exceeded present-day 
Europeans, the average for males being nearly six feet. The forehead was high; 
brow ridges were practically absent; the thigh bones were straight and the lower limbs 
long, indicating swift-footedness. The cranial capacity was above the average of 
today, perhaps as high as 1800 cc., even the women exceeding the average of present- 
day males. The chief peculiarity of the race was the combination of broad cheek 
bones with a narrow skull, suggesting that the Cro-Magnons somewhat resembled 
the modern Eskimo. 

These people were still in the paleolithic culture stage, for they used beautifully 
chipped but unpolished stone implements. They show great progress beyond the 
Neanderthals, however, in that they not only made use of bone, as well as flint, but 
carved it, and the carvings, drawings, and paintings ascribed to them are the wonders 
of archaeological discovery. Many of the caves of central France and the northern 
Pyrenees bear upon their walls drawings and paintings of hairy mammoths, steppe 
horses, bisons, reindeer, and other subjects. Plastic art is represented by carvings in 
bone and soft stone and, more rarely, by statuettes in clay. This art continued over 
three culture periods known as the Aurignacian, Solutrean, and Magdalenian. It 
had its beginnings in Aurignacian times near the close of the glacial period with 
figures incised on bone or on the walls of caves. These pictures are of contemporary 
animals, crudely drawn, the proportions poor, quadrupeds usually shown with two 
legs only. Considerable progress was made during Aurignacian times, but during the 
Solutrean art was more or less in abeyance. The Magdalenian period saw the cul- 
mination of prehistoric art, winning for the Cro-Magnons then living the distinction 
of being called the "Greeks of the Stone Age." Not only were animals depicted in 
their proper proportions, but lifelike poses were attained, and there was some success 
in composition. Colors were employed, and a few of the better-preserved paintings 
show four shades, obtained by the employment of red and yellow ochres and man- 
ganese with grease as a medium. 

East of the domains of the tall Cro-Magnons, on the plains of Moravia, lived the 
Briinn or Pfedmost race. What they knew of art or culture we cannot tell, for they 
decorated no caves. Many of the remains so far found are burials, deep in the regional 


loess. One great pit has produced fourteen skeletons, with fragments of six others. 
Above and below them were fragments of almost a thousand mammoths, a circum- 
stance which has led to the designation of this race as the "mammoth hunters." With 
the elephant bones were found implements of the "laurel leaf" type which show con- 
temporaneity with the mid-age of the Cro-Magnons. Other stations have produced 
implements of a simpler type, Aurignacian or older. The Predmost people differed 
from the Cro-Magnons in being somewhat less tall, about 5 feet 6 inches to 5 feet 
7 inches in mean height, in being heavier-boned, and in having a slightly smaller 
cranial capacity, only about 100 cc. more than that of modern Europeans. Their 
jaws were strong and somewhat prognathous, but the chin was prominent, not retreat- 
ing. A neanderthaloid feature is seen in the brow ridges of the males. Some anthro- 
pologists believe that this indicates that they may have been connecting links between 
Neanderthals and sapient men, but similar ridges are shown by some modern 
Australian bushmen. 

Another race, the Grimaldi, may have lived along the northern shores of the 
Mediterranean in early Aurignacian times. Much has been written about two un- 
usually well-preserved skeletons, mother and son, found in the Grotte des Enfants 
on the Riviera in 1906. The specimens have certain negroid characteristics which 
have aroused great interest. Were they connecting links between the Negroes and 
Caucasians? Or were they wanderers who somehow crossed the Mare Interius and 
through some service became entitled to honorable burial? Sir Arthur Keith points 
out in a recent book that not only the Grimaldi but some of the Predmost people 
showed negroid features. Such seem, in fact, to be characteristic of the whole Cro- 
Magnon stock, using that term in a broad sense for all the peoples who employed 
the Aurignacian, Solutrean, and Magdalenian types of implements. Keith is in- 
clined to think that this indicates that Cro-Magnons and negroids had the same 

With the first appearance of Homo sapiens, a nonspecialist must stop and let the 
anthropologist take over, even though the field of human paleontology is not neces- 
sarily so limited. 

It is not possible in a chapter of this sort to describe all known human fossils, 
but a few others will be mentioned in the discussion which follows. We are interested 
chiefly in seeing what light the present meager evidence throws upon the ancestry 
of man. Although there are several possible interpretations, the current theories fall 
under two headings. One view is that man is closely allied to the Simiidae and be- 
came separated from that group relatively recently, perhaps during the late Miocene 
or Pliocene. This is the opinion championed particularly by Professor William K. 
Gregory, who reached it after an intensive study of the teeth of all modern and extinct 
primates. Another is that man came from the same ancestors as the anthropoids, 
the group including monkeys as well as the Simiidae. In concrete terms, this would 


mean a parting of the ways as early as the Eocene, the lemuroids of that age being 
the common ancestors. This is the view most acceptable to conservative people, 
especially those who do not want a monkey for an ancestor. Few scientists, except, 
perhaps, those in their later years, accept this view in its extreme form, but there are 
many who believe that the Hominidae and the Simiidae separated from their com- 
mon ancestor in late Oligocene or early Miocene times. 

That man early left the simian line is suggested by the large brain and, par- 
ticularly, by the great toe, which has lost its primitive divergence from the others, 
become enlarged, elongated, and the principal bearer of the weight. Sir Ray Lankes- 
ter showed that this feature indicated that man's ancestors must have left their arboreal 
homes and become fully terrestrial many eons ago. 

Another fact which suggests an early divergence of the human and simian lines 
is that the young of modern apes are more manlike in appearance than the adults, 
whereas human babies are more apelike than fully grown individuals. In other words, 
human and simian infants are much alike but during the youthful stages continually 
diverge. Interpreted on the basis of the theory of recapitulation, this would mean 
descent from a common ancestry, not one from the other. Young apes have a 
big brain, a highly vaulted skull, a relatively high forehead, and no brow ridges. 
H. Klaatsch, and many others since, have inferred from this that the ancestral simian, 
although prognathous, was considerably more human-looking than the modern 
gorilla or chimpanzee. It is possible that from a common ancestor there were two 
lines of evolution: one in which the brain case became elongate, with heavy super- 
ciliary and in some cases sagittal crests, and in which prognathism increased rather 
than diminished; another in which the brain case increased in size laterally and 
upward, and the jaw retreated, producing a straight face. Since this idea involves 
the "biogenic" law, it has, naturally, been severely criticized. One critic has said 
that to suppose that the remote common ancestor of man and anthropoids had a 
vertical forehead without brow ridges is to invent an entirely hypothetical group; 
but in another place the same author has written of Parapithecus: "Although only 
the lower jaw is known, this highly important form must have had the shortened 
face and swollen brain-case, and probably the large eyes of the small insectivorous 
tarsioids." The deduction that the ancestor was a short-headed, large-brained, smooth- 
browed, weak-jawed creature seems to be sustained. 

The chief reason for believing that the split did not come till after Lower Oligo- 
cene is that this was the time of the first appearance of primates with short heads, 
the dental formula of apes and men, and lower molars with five cusps. The indica- 
tions are that Parapithecus and Propliopithecus sprang from some as yet unknown 
tarsioid stem, animals with large brains, large and forward-directed eyes, and small 
canines. There is no evidence of the intervention of a "monkey" stage, in the technical 
sense. Parapithecus or Propliopithecus, or both, are generally accepted as ancestors 


of both men and apes. It is unfortunate that we know so little about them* Several 
lines of descent from one or both of them are theoretically possible. One through 
the Pliocene Pliopithecus to the modern gibbons has not been questioned. A second 
through the Upper Miocene Palaeosimia to the Pliocene and recent orangoutangs is 
reasonable. A third through the Mid-Miocene Dryopithecus and Pliocene Paleopithe- 
cus to the gorilla and a fourth through Miocene and Pliocene species of Dryopithecus 
to the chimpanzee seem probable. A fifth, through some unknown species of Dryo- 
pithecus to Homo sapiens, has been strongly advocated. One route (53) is through 






Gibbons Orangs Gorilla Chimpanzee 

Pliopithecus \ Paleo- 

\ Simia pithecus 


^Paleosimia \ S 


/ H neander- I Eoanth- H. neander- \ 
/ thalensis ropus thalensis 
Eoanth- Pith , an _ v J 

ropus thropus I ySinanthropus 

Sinanthropus \ Australopithecus 

Unknown Tarsioid 

FIG. 149. Diagram showing some of the lines of descent of modern primates 
which are theoretically possible. It should not be interpreted as suggesting that 
Homo sapiens is polyphyletic. 

Eoanthropus with enlarged canines and a high forehead, the other (5b) through his 
low-browed, small-canined cousins, Sinanthropus, Pithecanthropus, Paleanthropus> 
and Homo neanderthalensis. Opinions differ as to whether all these Pleistocene 
men are in the direct line or whether some or all of them represent lateral branches, 
A sixth possible line has three variants: (6a) from some tarsioid through unknown 
descendants with enlarged brains, smooth brows, and decreasing canines to Homo 
sapiens; (6b) the same, with enlargement of canines, to an Australopithecus-like, ape 
stage, thence to Eoanthropus, and, by decrease in canines, to Homo sapiens; (6c) the 
same, with reduction of canines, through the superciliarily ridged Sinanthropus- 
Neanderthal line, to Homo sapiens] and (7) a line from Parapithecus, in which the 
canines are no larger than the adjacent teeth, to modern man, with no known con- 
necting links. It is obvious that it is impossible to discuss these lines in any detail. 


They are listed merely to show how wide a range of speculation the present in- 
adequate evidence permits. 

The chief objection to the derivation of man from Dryopithecus (line 5) lies in 
the fact that all known species of that genus and they range from Mid-Miocene 
to Pliocene have large canines, and there is no indication within the genus of any 
reduction (Fig. 147 B). In fact, the whole trend seems toward increase in size, as is 
shown by the two known descendants, the chimpanzee and the gorilla. Much has 
been made of the fact that man and all the post-Miocene simians show the Dryo- 
pithecus pattern in the molars. Yet it is generally admitted that the orangoutangs and 
the gibbons are not descendants of Dryopithecus but each the result of an independent 
line of evolution from Propliopithecus. Hence the "Dryopithecus' pattern is really 
an inheritance from the Oligocene ancestor, and the argument loses much of its 
weight. The presence of Dryopithecits-likc molars in man is no proof of descent from 
that genus, for he, like the orang and the gibbon, may have inherited them from 
some Propliopithecus-likc ancestor. 

There has been a general tendency to treat the subject of the reduction of the 
canines rather lightly, as though it were a thing which might happen almost at 
will. Some see no reason why Eoanthropus (line 53) of the Mid-Pleistocene should 
not have evolved into Homo sapiens before the end of the glacial period. Perhaps 
he might, if he had suddenly changed his disposition and gone on a diet of poached 
eggs and milk toast. But it is doubtful. Remember how long it took the horse to 
get rid of the extra toes. Once a trend is established and inheritance gets hold of it, 
a change to an opposite one is difficult. Atrophy can, so far as is known, come about 
only through disuse; and it is difficult to disuse teeth, as all know who have had 
occasion to "favor" some sensitive one. Still more difficult to accept is the opinion 
that canine reduction has passed through two cycles in the history of the primates. 
Paleocene and Eocene lemuroids with large canines led to Eocene tarsioids, as yet 
unknown, and they to the Lower Oligocene catarrhincs. From its reduced condition 
in Propliopithecus and Parapithecus the canine was gradually built up to a fang 
again, the status it had in Eocene times. Then, after the Miocene, it dwindled away 
in the line which led to man. Although such a history is not impossible, it is im- 
probable. It would be difficult to find any parallel in the animal kingdom. 

The' heavy brows of the Pithecanthropus-Neanderthal series (line 5b) have been 
the chief stumblingblock in the way of their acceptance as ancestors of modern man. 
It is difficult to imagine so radical a change as that which would be necessary to 
transform Homo neanderthalensis into Homo sapiens. Furthermore, there is little 
doubt that the two races were contemporaneous for a time at least. If Homo sapiens 
is a descendant of a heavy-browed race it must have been a pre-Neanderthal one. 
Perhaps it was Paleanthropus, as Gregory has suggested. It may be that Neanderthal 
man was derived from an ancestor less brutal-looking than himself, for both the 


Galilee and the Ehringsdorf skulls are more highly vaulted than those of the typical 
members of their race. There is considerable evidence that Ehringsdorf man may 
have invaded Europe during the warm period which predated the last (Wurm) 
glaciation, and that, although the Galilean neanderthaloids lived during glacial 
times, their home in Arabia may then have had a moist and comfortable climate. 
Perhaps the rigors of arctic conditions served to brutalize the later members of the 
race, though this seems hardly probable, for Homo rhodesiensis, who escaped to 
South Africa, has the strongest brow ridges of all. 

On the other hand, some Neanderthal children (Gibraltar and La Quina 
finds) afford the same sort of evidence as the skulls of young simians do. Brow 
ridges are absent, and the forehead is moderately high. If this is an inherited and 
not an adaptive characteristic, it can be interpreted as pointing toward a smooth- 
browed ancestor for the neanderthaloids. 

It is unfortunate that not a single skull or even a calvarium of a Tertiary simian 
has so far been found. Even Dryopithccus is known from jaws only. It is commonly 
thought that the Miocene and Pliocene members of this group must have had 
heavy brow ridges, because we naturally infer from the structure of the higher apes 
that brow ridges and enlarged canines go together, a Cuvierian deduction. But this 
is not true, for some of the larger-brained South American Cebidae have smooth 
foreheads and large canines, and the hominids just discussed have brow ridges and 
small canines. 

There can be no question but that everyone in any way connected with the 
problems of human origin would have been much happier if Piltdown man, Eoan- 
thropus dawsoni, had never been discovered. As has already been pointed out, there 
is small chance that he was ancestral to modern man; yet he has too many human 
characteristics to be excluded from the family Hominidae. If he is a descendant of 
Dryopithecus, as seems probable, and if Homo sapiens and the Pithecanthropus- 
Neanderthal clan are not, as also seems probable, then the family is polyphyletic, 
and a large-brained man has been evolved independently along two (or more) 
different lines. This would not disturb us if we were dealing with mere "animals," 
for no one objects to the derivation of the orangs, the gibbons, and the chimpanzee- 
gorilla group over three different routes. But the idea that big-brained man, supreme 
among all earthly beings, should have been evolved twice seems repugnant to scientist 
and layman alike. No wonder that American paleontologists objected to Smith 
Woodward's restoration, transferred the canine from the lower to the upper jaw, 
and made the teeth the type of a new species of chimpanzee! 

The description by Professor Raymond Dart of what has come to be known as 
the Taungs skull is of interest in this connection. It was found in a filled cave near 
Taungs in Bechuanaland, South Africa, where it was associated with remains of other 
mammals, including three extinct species of baboons. The age of the deposit is un- 


known, but it was probably formed during the Pleistocene. When the matrix had 
been removed from the specimen, there remained a part of the skull retaining the 
endocranial cast, the face, and the more important parts of the jaws. The teeth 
showed it to be a young individual, and detailed study of the cast of the cranial 
cavity proves that it was a simian, allied to the gorilla and chimpanzee. But its 
brain capacity, 500 cc., is far above that of any of its relatives when at the same age. 
It also shows such human features as a high forehead, lack of orbital ridges, a para- 
bolic dental arch, small canines, no diastema in the lower jaw, no simian shelf, and 
an advanced position of the foramen magnum, which suggests an upright position 
(either sitting or standing) . As Sir Arthur Keith says, if this skull had been found 
in a Miocene deposit, it would have been seized upon as an excellent connecting 
link between simian and human lines. Since it is probably not older than Pleistocene 
it has to be interpreted as the young of some ape, allied to the gorilla, but with larger 
brain and other human attributes. Since this paragraph was written, Dr. Broom has 
found an adult in a near-by region. It is much like a chimpanzee, but the teeth 
have human characteristics. 

To a certain extent, a dual ancestry of Pleistocene man brings us back to 
Klaatsch's now abandoned theory of two derivations, one from light-boned Asiatic 
ancestors, the other from heavy-boned Africans. The detailed phylogenies would, 
however, be entirely different. It is futile, in the present state of knowledge, to con- 
tinue this discussion. It has been carried thus far only to disturb those who think 
that the problem of the ancestry of man has been solved. 

If we recapitulate briefly what has already been said, it appears that the primates 
became differentiated from the insectivores in the Paleocene and that some among 
the late Eocene tarsioids showed a beginning of large cranial capacity. From such 
stock the Anthropoidea arose in Egypt during Oligocene times. The particular 
forms which are looked upon as ancestral (Propliofithecus and Parafithecui) are 
small, monkeylike creatures. From the first, primates were arboreal in their habits 
and naturally tended to become fruit and nut eaters, although the early forms were 
more or less omnivorous and, throughout their history, were ready to add insects, 
lizards, and other small animals to their diet. The Eocene and Oligocene were in 
Asia and Africa times of moist and equable climates, and forests were extensive. 
Under these favorable conditions, the primates increased in size, and, as is known 
from the considerable number of species found in the Miocene of the Siwalik hills 
in northern India, Asia became a region in which they were abundant. In that 
locality lived the various species of Dryopithecus, the ancestors of the chimpanzee 
and gorilla. 

During the Miocene, the Himalayas and the mountains of southern Europe began 
to rise, obliterating the Asiatic part of the ancient Mediterranean; consequently the 
climate changed, and the present aridity of central Asia was gradually acquired. 


Barrell has suggested that this would cause the thinning and final disappearance 
of the forests from the region north of the Himalayas and that, as a result, any 
pre-hominids in that region would gradually be forced from the ancestral arboreal 
life to existence on the ground, with the consequent change to an upright position, 
which arboreal life had made familiar. Increased aridity meant not only loss of 
trees byut also a decreased supply of food, so that man could no longer depend largely 
on fruit. He became a hunter, and naturally increased in cunning and swiftness. 
The use of the teeth changed with the nature of the food. In the Neanderthals and 
their Heidelbergian ancestors we see the development of the "edge to edge bite," 
which would be well fitted for gnawing meat, and of enormous jaws with strength 
enough to crack bones as well. Further, as men on the ground came to use weapons 
rather than their teeth in fighting, so the canines, if originally large, may have 
dwindled away. Those anthropoids which were on the southern slopes of the moun- 
tains were not deprived of forest conditions. Consequently, they retained their an- 
cient habits of gathering food and fighting with the teeth, and evolved in an entirely 
different direction from that of their cousins farther north. Fossils to support this 
theory of BarrelPs have yet to be found. Nevertheless it may serve as a working 
hypothesis in the search for more evidence. 


Who can see the green Earth any more, 
As she was by the sources of Time? 

Matthew Arnold, "The Future" 

Was the earth green "by the sources of Time"? Neither geologists nor paleo- 
botanists can answer the question. Probably the continents were dreary, unclothed 
wastes without visible inhabitants at least till late Cambrian times, possibly even 
later. All that has been learned from the study of their morphology, and particularly 
from their methods of fertilization, indicates that terrestrial plants, like terrestrial 
animals, had aquatic ancestors. This inference is borne out by the record, for few 
have been found in strata older than the Lower Devonian. It may be that the record 
is faulty, since no strata which can be proved to be non-marine have yet been found 
in formations deposited before late Silurian times. Curiously enough, the oldest 
supposedly estuarine or fresh-water beds contain no terrestrial plants. Woody tissue 
is characteristic of vegetation on land; until it was formed there was little chance 
for preservation of such organisms. Since cellulose-bearing fossils occur in no strata, 
either fresh-water or marine, before the Upper Silurian, it seems probable that there 
was little "green earth" before that time. 

The known pre-Devonian floras are essentially aquatic. Only such simple 
forms as blue-green, green, red, and brown algae are actually represented by fossils, 
but the existence of bacteria and diatoms can be inferred from what appears to be 
compulsive evidence. 

Bacteria are animal-like plants, or plantlike animals : their status in one's estima- 
tion depends upon one's point of view. They lack the chlorophyll of plants, and 
most of them feed upon organic matter; hence, by definition, they are animals. 
But some, despite their lack of green coloring matter, are able to abstract carbon and 
oxygen- from carbon dioxide and its compounds, and so should be classed as plants. 
Some of them are probably but little changed from the hypothetical plant-animals 
which are supposed to have been the first living creatures on this planet. Simple as 
they are in structure, however, bacteria, like all other organisms, have changed 
(evolved) in the course of time. Many, unfortunately, have become parasitic, their 
activities resulting in diseases of various plants and animals. Since these affect the 
human individual directly, the group is more or less in disgrace, and few people 
seem to realize that the majority of bacteria are beneficent. They are the chief 


agents in the formation of soils, for without the acids and alkalies they produce the 
decomposition of rocks would be an exceedingly slow process. They alone among 
organisms have the power of taking nitrogen from its solution in water and building 
it into compounds available to plants. They are the agents of putrefaction and fer- 
mentation. Without them, agriculture would be impossible; in fact, one might go 
so far las to say that, were it not for them, terrestrial vegetation would never have 
become abundant. Sediments of pre-Cambrian age show the same physical charac- 
teristics as those of later times; hence it may be inferred that bacteria have existed 
since the earliest periods of earth history. The oldest actual records of their presence 
are furnished by certain late Devonian ostracoderm plates in which the loss of the 
original structure appears to be due to these organisms. 

Closely allied to the bacteria are the unicellular blue-green algae. Their principal 
importance lies in the secretion of calcium carbonate, and their early history has already 
been discussed in the chapter on the Pre-Cambrian. Of the modern seaweeds the 
most conspicuous and most highly organized are the kelp, the brown algae. They 
are restricted almost entirely to marine situations, growing abundantly on rocky 
coasts. Like other algae, they have no roots or leaves, and take their food from 
the dissolved gases and chemical compounds in the medium in which they live. 
Probably they have been in existence since the Pre-Cambrian, but as they form 
neither woody tissue nor calcareous deposits they are ill adapted for preservation, and 
their history is obscure. More important from a geological standpoint are the green 
algae, forms in which the color of the green chlorophyll is not obscured by other 
pigments. They live in both salt and fresh water, are less specialized than their 
brown relatives, and have two claims upon our interest. Some secrete calcium car- 
bonate, contributing greatly to the substance of coral reefs; others form the marl of 
inland lakes and bogs, a substance much used by farmers to "lime" their lands. 
Still others, more simple, were the probable ancestors of terrestrial plants. As for 
the red algae, everyone who visits the seashore knows the pink or white encrustations 
they form on shells and rocks. Their contribution to modern coral reefs is great; 
in fact, it is doubtful if, without the protecting and binding qualities of Litho- 
thamnion, corals would be able to build reefs. Since these algae cause the deposition 
of both calcium and magnesium carbonates, they have also contributed greatly to 
the formation of limestone and dolomite. 

Whatever their good qualities, algae are not important in supplying food. That 
role is now played by the diatoms, minute unicellular plants which are the funda- 
mental source of food for marine animals. Although they are microscopic in size, 
even the national debt seems small as compared with the number present in a quart 
of sea water during the reproductive season. Innumerable billions of them are present 
at all times in the upper layers of the oceans, particularly in cold water, and during 
the so-called "flowering" periods, from March to September, they are so abundant 


that the water has been called a "vegetable soup." Fortunately they are present in 
fresh water as well as in the sea, for all small aquatic animals are primarily dependent 
upon them. With them in surface waters are protozoans and the larvae of all sorts 
of animals, both benthonic and nectonic, and it is the fate of these minute plants to 
serve as food for their associates. Since many of the tiny animals are in their turn 
eaten by larger ones, there are regular cycles in the building up of larger and larger 
creatures. Small planktonic crustaceans, copepods, eat the diatoms; small fishes live 
on the crustaceans; larger fishes eat the small ones. (Naturally, the larger the in- 
dividuals, the fewer their numbers.) This cycle is but one of many. Although they 
are particularly abundant in the plankton, diatoms are also numerous on the floors 
of seas and lakes, and on the surfaces of aquatic plants, where mollusks, worms, 
and other benthonic animals find them. 

Unfortunately, the geological record of this group is incomplete. There are 
doubtful records from the Paleozoic; not till the Jurassic are there deposits with 
identifiable specimens. But there are two good reasons why this should be so. In 
the first place, the plants are extremely minute, mensurable only in fractions of 
millimeters; in the second, their skeletons are exceedingly fragile. Their tests, unlike 
those of other plants, are siliceous and consist of two shallow "pans," one of which 
fits into the other like a pillbox into its lid. Even the lightest pressure reduces them 
to powder of unrecognizable components, so fine that some of the most delicate 
polishing media are made from diatomaceous deposits of Tertiary age. 

Just as there are two explanations for the failure of these plants to be preserved 
as fossils in the older rocks, so there are two reasons for believing that they were 
in existence even in pre-Cambrian times. The more obvious is that there were pre- 
Cambrian animals, which must have had vegetable food. Calcareous algae were too 
well protected to furnish it in any quantity; brown algae were too large and tough, 
and too completely confined to the strand to be of any great use to the minute animals 
of the plankton. Hence, diatoms, or their naked ancestors, must have been present. 
Secondly, it will be remembered that our discussion of the conditions of early pre- 
Cambrian time led to the suggestion that the original oceans were somewhat acid, 
compelling the earliest animals to secrete chitinous or siliceous skeletons, if any. 
Perhaps the diatoms began to form hard parts at the same time as the siliceous 
sponges and radiolarians. 

Mr. William C. Darrah has recently discovered spores of terrestrial plants in 
Swedish Upper Cambrian shales, but the nature of the vegetative shoots which bore 
them is as yet unknown. The oldest recognizable stems are found associated with 
graptolites in the Upper Silurian of Victoria, Australia. They are similar to forms 
which have long been known from Lower and Mid-Devonian strata in North 
America and Europe. All are small spore-bearers, and hence allied to the ferns, 
although they have a simpler foliage, leaves being absent or short and linear. The 


roots are poorly developed, but the underground stems are extensive (Fig. 150). 
Such plants as these were probably in existence as early as Ordovician times, for they 
fit in well as connecting links between aquatic and terrestrial types. The former 
needed no roots or leaves, for they could take water, oxygen, carbon dioxide, and 
other food directly from the medium in which they lived. The latter depend on 
their roots for water and foods dissolved in it, and upon their leaves for "breathing" 
and for the transformation of raw materials into food. 

Professor Douglas Campbell has developed the most satisfactory theory of the 
evolution of terrestrial from marine plants. He thinks that the ancestors were 
green algae, for they alone lack other sorts of pigment, and that their first migration 
was from the sea into fresh water. At times of low water in rivers and pools, the 
small weeds came into contact with the mud. The first reaction probably was 
longitudinal growth; the second, the protrusion of delicate, hair-sized rootlets to 
keep connection with water beneath the surface. Finally, in response to the influence 
of light, upright shoots were formed. The production of the first land plants was 
so important an event that the processes involved seem wonderful and mysterious. Yet 
the results, and probably the processes, were less remarkable than those physiological- 
chemical reactions which take place in our gardens every season. The really extraor- 
dinary thing is that new terrestrial forms are not constantly being evolved from 
simple green algae. 

The situation is parallel to that of the evolution of terrestrial animals. Probably 
plants, snails, insects, arachnids, and amphibians emerged from the aquatic environ- 
ment at about the same time. Only in the case of the amphibians is the record suffi- 
ciently full to allow reasonable inferences to be drawn. In that case the result seems 
to have been due to the fact that certgjn geographical changes happened to coincide 
with a particular stage of the evolution of a particular group of fishes. Changing 
physical conditions repeat themselves at intervals of varying lengths, but evolution 
in organisms is a cumulative process; hence there can be no exact repetition. No 
matter how many times particular environmental conditions are repeated, there is 
only one time in the history of any group of organisms at which it can respond to 
external influences in one particular way. 

The state of evolution of the late Silurian and early Devonian flora supports 
Campbell's theory. The objection has been raised that the nearly rootless condition 
of these early forms was a secondary xerophytic adaptation, for it may be that they 
lived in bogs. The criticism is probably just, but the inference therefrom is mis- 
leading. Bogs, seasonally dry, are ideal locations for transitions from semiaquatic 
to semiterrestrial life. Characteristics which were primitive in the early days may 
appear later as secondary features. Roots may come and roots may go. Moreover, at 
least one genus prominent at this time, Psilophyton, is so widespread geographically 
that it seems unlikely that it was confined to swamps. 


The Mid-Devonian witnessed one of those sudden influxes of diversified plants 
which have led to the idea of "explosive" evolution. After countless ages terrestrial 
floras came suddenly upon the scene. There really is no great mystery here, merely 
lack of information. When more pre-Devonian terrestrial deposits are found, their 
fossils will reveal more about the evolution of plants. 

Ferns have been in existence longer than any other plants familiar to us today. 
After their modest beginnings in the Mid-Devonian, their progress was rapid, reach- 
ing its culmination in the Carboniferous. Throughout the late Paleozoic and Meso- 
zoic they formed the undergrowth in moist, shady places, just as their descendants 
do today. With favorable conditions tree ferns have from time to time risen above 
the general level, but they have never been dominant. Their beautiful foliage makes 
ferns popular favorites, but quite as fascinating is their Paleozoic habit of spreading 
by means of underground stems. To this, and to the fact that they seem never to 
have been particularly acceptable as food, they probably owe their survival. 

The great trees of the late Paleozoic forests (Fig. 151) were also sporebearers, 
relatives of the modern ground pines and horsetail rushes, now comparatively in- 
significant plants. Most conspicuous were the lepidodendrons and sigillarias, with 
straight stems reaching heights of a hundred feet or more. Lepidodendron branched 
freely and had short, grasslike leaves and a trunk ornamented by diamond-shaped, 
diagonally-arranged leaf scars. Sigillaria had few branches, much longer leaves, and 
leaf scars in vertical rows. The fructifications of both included large, almost seed- 
like megaspores, more highly organized than those of their later relatives. Both 
trees spread laterally by means of bifurcating underground stems, commonly called 
Stigmaria. These served not only for the production of new trunks, as in the ferns, 
but also for the support of the trees. This facl and the relatively small size of the 
roots indicate that the soil in which they grew was swampy. Catamites was the late 
Paleozoic representative of the rushes. The vertical fluting of the trunk, its nodes, 
from which whorls of branches emerge, and its radially arranged leaves are so like 
those of modern forms that no one would fail to recognize its affinities. The leaves, 
known as Annularia, are longer and more graceful than those of living species, and 
the height attained by some, sixty or seventy feet, is not reached even by modern 
tropical representatives of the group. 

Although spore-bearers dominated the late Paleozoic vegetation, seed plants ap- 
peared as early as the Mid-Devonian. The oldest are those with fernlike leaves, the 
pteridosperms (Fig. 152). Since they have the foliage of their putative ancestors, 
the spore-bearing ferns, they are believed to be the primitive members of their group. 
During the Carboniferous they were abundant, but it was their fate to disappear at 
the end of the Paleozoic. There is considerable evidence to show that they were 
ancestors of various other seed-producing groups. Some were vinelike in habit, 
others similar to tree ferns. It is probable that they produced many of the nutlike 

FIG. 150. At left, Rhynia, at right, Astcroxylon, primitive Mid-Devonian 
spore-bearers. Note the absence of leaves from the former, their primitive 
condition on the latter. From Kidston and Lang. 

FIG. 151. Resloraiiuii ui a Carboniferous rorest, showing Lcpidodcndron, 
Sigillaria, ferns, and fernlike plants. Photograph by courtesy of the Field 
Museum of Natural History, Chicago. 

FIG. 152. A pteridosperm, Emplcctoptcris, with seeds in place upon the 
pinnules. An original drawing, contributed by Mr. William C. Darrah. 

FIG. 153. At left, casts of three hickory-nut meats from the Oligocene of 
Nebraska. At right, a coconut from the Eocene of Belgium. Photograph of 
the latter through the courtesy of W. C. Darrah. 


seeds so common in Carboniferous rocks, and for long a puzzle to paleobotanists. 

True gymnosperms, with naked seeds and simple foliage, also came on the 
scene during the Devonian. The earliest, and the most abundant during the later 
part of the Paleozoic, had long, strap-shaped leaves with a parallel venation similar 
to that of the modern maize. Best known of these is Cordattes y commonly called the 
first "fruit tree," for its yew-like seeds appear to have been enclosed in a fleshy capsule. 
Short-leaved conifers, much like modern evergreens, did not appear until the Permian. 

Paleozoic terrestrial floras were of geological importance in producing ancestors 
for later ones and in supplying materials for the formation of coal. But as a source 
of food they were negligible. Terrestrial animals starved in a land of plenty. Am- 
phibians inherited their food habits, along with their teeth, from carnivorous fishes 
and transmitted both to the early reptiles. In Carboniferous times the food cycle, even 
of the air-breathers, remained much as it had been in the olden days. Small inverte- 
brates and fish ate the unicellular plants in the rivers, small amphibians and reptiles 
ate the small invertebrates and fish, and larger fish and tetrapods fed upon all sorts 
of defenseless creatures. The minute plants in the water were still the primary source 
of food. A strictly carnivorous population, on land, has its limitations. So long 
as its primary source was in the fresh waters food was scarce, a fact which probably 
accounts for the small size of most Paleozoic amphibians and reptiles. A few giants 
were seven or eight feet long, but the average size was nearer seven or eight inches. 
All were semiaquatic, and dwelt beside the sluggish rivers on the coastal plains bor- 
dering the seas. Probably, in season, they came upon seeds or the tasty fruits of 
Cordaites, and they may have gulped some of them. Possibly they even cast specu- 
lative eyes on the source of supply. But they were not built for climbing; it was as 
much as they could do to drag themselves along on their bellies. They had no real 
taste for plant food. Millions of years were to pass, and a much more attractive table 
was to be set, before vertebrates really became vegetarians. Few reptiles and almost 
no amphibians ever became reconciled to a diet of vegetables. They would rather 
die than eat the stuff. 

Nevertheless, the increasing amount of vegetation began to make its influence 
felt, although indirectly at first. Decaying plants furnished food for various terrestrial 
arthropods and their larvae that is, for diplopods, centipedes, and many insects, 
all groups which first appeared or which first became abundant during the later part 
of the Carboniferous. Their presence probably first turned the attention of land 
animals away from the water, and grubbing for them may have provided the first 
vegetarian dinner. 

The world-wide uplift of the continents at the end of the Paleozoic drained the 
Carboniferous swamps and extinguished the gigantic spore-bearers and the seed- 
bearing fernlike plants. Increase in the size of the lands gave new opportunities for 
the floras to expand, but under new conditions. Plants had to dig for water. It was 


a time of development of root systems rather than of upright shoots. In a sense, 
plants had to start over again. As a result the floras of the Triassic and Jurassic were 
not particularly luxuriant and contained few lofty trees. They have been described as 
"scrub." Tree ferns, small evergreens, monkey puzzles, ginkgoes, and cycadoids domi- 
nated the scene, with ferns and small lycopods in the undergrowth. 

This was not only the real age of gymnosperms but the time when Yeptiles got 
their feet solidly planted on dry land. Much has been written about the amnion 
and the allantois, one essentially a water-cushion for the embyro, the other inti- 
mately connected with the shell of the egg, providing a means of respiration in the 
unhatched state. The presence or absence of such structures is said to be a funda- 
mental difference between reptiles and amphibians. But was it in the early days? 
The geological record suggests that the early reptiles, like Amphibia, laid their eggs 
in water, and that somehow, as the lands became dryer, there gradually evolved 
in the reptiles a type of egg which could be hatched on land. 

"Propinquity leads to love and marriage." Why not to changes in food habits? 
Vegetable food was plentiful in early Mesozoic times not only along the streams but 
on the uplands. Reptiles gradually acquired a taste for it, and throve; the old- 
fashioned stegocephalians stuck to the rivers, and perished. Unfortunately, vege- 
tarianism did not help vegetarians much in Mesozoic times; they fattened themselves 
merely to fall prey to the carnivores. Not until the Cretaceous did the herbivorous 
dinosaurs become abundant. Their conspicuous evolution coincides with the sudden 
appearance in North America, Europe, and New Zealand of the first plants with 
conspicuous flowers, the angiosperms. 

These plants, dominant today, made a spectacular entry upon the world stage 
as great trees rather than as modest herbs, and paleobotanists are still ransacking 
the earth in search of their ancestors. The oldest known are singularly like forms 
now living. Their wood, leaves, flowers, and fructifications are of modern type. The 
two great groups, monocotyledons, with parallel venation (palms), and dicotyledons, 
with reticulate venation (common hardwoods), had already been differentiated. At 
one time or another during the Cretaceous, forests much like modern ones were estab- 
lished. Many of the trees were sorts familiar today. Willows, birches, magnolias, 
tulips, sycamores, cottonwoods, sassafrasses, and viburnums were common, but more 
important sources of food were the figs, persimmons, breadfruits, oaks, beeches, 
walnuts, and palms. This rapid increase in the supply of seeds, fruits, nuts (Fig. 153), 
and edible leaves had a profound influence upon the evolution of arboreal reptiles, 
birds, and mammals, as has been related in other chapters. 

The problem of the herbivorous dinosaurs requires a moment's further con- 
sideration. Their first abundance coincided with the arrival of the angiosperms. 
But is there really any connection between the two events? As has been pointed out, 
there was a noticeable lag in the adaptation of reptiles to vegetable food. It is prob- 


able that not till about the end of the Jurassic did they become used to feeding upon 
the various gymnosperms surrounding them. Once they were so accustomed, their 
opportunity had arrived, and they increased and multiplied. Their bills and cheek 
teeth were well fitted for plucking and slicing the coarse trunks and leaves of the 
cycadoids, plants which reached their maximum in the late Jurassic and early Cre- 
taceous, and then gradually faded from the scene as the angiosperms replaced them 
in late Cretaceous and early Tertiary times. It may be that the herbivorous dinosaurs 
were so restricted in their diet that the coming of the flowering plants was no advan- 
tage to them. It is even possible that it caused their extinction. 

The Paleocene flora much resembled that of the late Cretaceous, though it in- 
cluded such new arrivals as the maples, with winged seeds, and, more important, 
the grasses. The Oligocene and Miocene, however, witnessed geographic changes in 
various parts of the world. Gradual uplift brought the Great Plains of western North 
America to high altitudes, and concomitant erosion of the rejuvenated Rockies flooded 
them with sediments which raised their surfaces still further. The two processes united 
to destroy forests and to produce a vast area occupied by grasses and herbs. Still later, 
orogeny brought up the coast ranges which now catch most of the Pacific moisture 
on their western slopes. This reduced the high plains to the status of a semiarid 
region and gave great opportunity for the spread of grasses. The effect of this on 
the evolution of mammals was all-important. 

The general uplift of continents which began in the Miocene culminated in the 
Pleistocene, a time when continental glaciation destroyed all vegetation over vast 
areas, later to be repopulated. Trees are not the first plants to occupy a previously 
devastated area. Ahead of them march a host of lowly angiosperms, grasses, and herbs, 
which can exist under unfavorable conditions. It is to them and their even more 
humble associates, bacteria, fungi, and ferns, that trees owe the preparation of the 
soils which permits them to advance. 

It is evident, in short, from all that we have said, that nature, as a food provider, 
has been somewhat slothful. All animal life is dependent upon plant food, but till 
Mesozoic times few if any vertebrates ate terrestrial vegetation. Larger animals fed 
upon smaller ones in a descending series to the smallest, which, perforce, consumed 
unicellular plants, for there was nothing else they could ingest. Carnivores dominated 
the world. Direct use of vegetable food, to which must be ascribed the manifold 
changes that have taken place in birds and mammals, did not begin until the Meso- 
zoic, and did not increase greatly until the late Cretaceous and Tertiary, when at 
last the angiosperms, the "life-givers," became the dominant plants. Humble servi- 
tors of the animal kingdom, they deserve an even deeper gratitude than that ex- 
pressed in the admiration aroused by their beauty. 


The Present is the living sum-total of the whole Past. 

Carlyle, "Characteristics" 

Carlyle's words are those of a writer who dealt with a limited period, but they 
are as true for the paleontologist for whom history began two billion years ago as 
they are for the historian for whom it began seven or eight thousand years ago. 
The viewpoints of the neo-historian and the paleo-historian differ widely. Carlyle 
and his predecessors, contemporaries, and successors give the impression that history 
began when man learned to write and that it ceased when they themselves laid 
down their pens. They cannot realize that the years which have been the subject 
of their study are so short a part of past time that the geologist has no unit in his 
calendar small enough to record them. Three or four thousand years seem a long 
time to one who has trouble finding out what happened a century ago. The paleon- 
tologist, like the astronomer, after years of study eventually acquires a sense of time 
which cannot be transmitted to others by the mere telling. He is therefore less prone 
than the ordinary historian to stress the present, and less apt to be disturbed by those 
contemporaneous happenings which seem unfavorable to the progress of civilization. 
Millions, not hundreds or even thousands, of years must pass before the fate of a race 
is determined. 

At the present moment man is but a few generations removed from the brute. 
That does not mean a few generations from the monkey, but a few from ancestors 
as tall, as big-brained, and as highly developed physically as ourselves. Self-preservation 
was not only the first law but practically the only law of man a few centuries ago, 
so few, in fact, that, geologically speaking, it was no time at all. Man is just beginning 
to learn from experience. Perhaps after a few hundred generations the accumulated 
knowledge of what has happened in the past may lead to a happier period for those 
who spend their brief lives on this planet. 

Homo sapiens appears to have been one of the latest productions of nature, 
for there is no record of his presence till near the close of the Pleistocene, seven to 
twenty thousand years ago a moment as compared with the supposed billion and 
a half years since the time when life may have appeared on the earth. Since man's 
span of existence probably will be comparable to that of other animals, he has a good 
"expectation of life." 

If one looks at the history of other groups one finds that, in general, after an 


inconspicuous beginning each experienced a period of great differentiation and ex- 
pansion, followed by a gradual or, more rarely, sudden decline in importance. The 
cystids furnish an excellent example. A narrow line would suffice to indicate the 
traces of their existence from the Lower Cambrian to the Lower Ordovician, the time 
when they suddenly became abundant; after the Mid-Ordovician they gradually 
declined in numbers till their extinction in the Mid-Devonian. Crinoids have the 
same history, although they first appeared in the Mid-Ordovician, reached their cul- 
mination in the early Mississippian, and have lingered on to the present. The old- 
fashioned tetracorals have a vague, almost traditional ancestry in the Mid-Cambrian* 
reached their high point in the Mid-Silurian and Mid-Devonian, and then faded 
from the scene in the Permian. Among the vertebrates the amphibians, after a feeble 
beginning in the Devonian, blossomed forth in the Permian, then declined to their 
present lowly status. Reduced to a diagram, the history of the reptiles shows the same 
figure, starting in the Pennsylvanian and culminating in the late Mesozoic, since when 
their variety has remained relatively small. The story of the mammals is similar, a 
poor start in the late Triassic leading to the extraordinary differentiation in the 
Tertiary, to be followed by the beginnings of restriction in late Pleistocene and Recent 

These examples have been drawn from large groups, but the same "law" holds 
true in lesser ones, whether they be orders, families, or genera. To cite examples, 
however, it would be necessary to descend to technicalities that would interest the 
specialist only. There are, in fact, so many such histories that paleontologists have 
come to accept them as examples of the usual course of events. Hence one infers 
that, since man ij> in the initial stages of his phylogeny, the probabilities are that his 
line will continue far into the future. 

As with all rules, there are exceptions, although not many. Those of one class 
are on the optimistic, those of the other on the pessimistic side, so far as predictions 
about the longevity of man are concerned. Examples of the former are the gastropods 
and pelecypods. Starting from scratch in the Lower Cambrian and Mid-Ordovician 
respectively, both groups have enjoyed continuing prosperity, reaching their greatest 
variety at the present day. It is possible that they have not yet attained the height of 
their evolution, even though man has eliminated many terrestrial forms in recent 
years. Insects show the same increasingly rapid upward swing; man's campaign 
against the more noxious of them has been only moderately successful. Such examples 
as these suggest the possibility of almost eternal life for man, but there are others 
which present a darker picture. One is that of the blastoids, whose brief story has 
already been recounted in the chapter dealing with the radiates. Their unimportant 
beginnings in the Ordovician and Silurian are comparable to those of the groups 
already discussed, but their time of extraordinary abundance in the late Mississippian 
was followed by almost complete extinction in the early Pennsylvanian. Several 


groups of crinoids ran against similar walls early in the Mississippian, and there are 
other orders or families for example, the ammonites and the ceratopsian dinosaurs 
which were extinguished at the height of their evolution. Man may have a simi- 
lar fate, but the chances are against it. , 

One of the principal reasons for so thinking is that, although man likes to believe 
that he sits on the topmost branch of the family tree, he really does not. As has already 
been shown, he is rather a primitive mammal, and if he has any regard for the future 
he should be glad of it. Edward Drinker Cope long ago pointed out in his "doc- 
trine of the unspecialized" that simple organisms appear to attain greater geological 
longevity than more elaborately constructed ones. Dr. Rudolf Ruedemann showed 
some years ago that forty-five Paleozoic genera have living species and that ten of 
them have survived since the Ordovician. All are invertebrates, and all are members 
of the more lowly classes: foraminifera, worms, brachiopods, ostracods, and the like. 
All have simple skeletons so simple, in fact, that it is a question whether the 
genera can be positively identified. It is probable that if the soft parts were known 
it would be found that those of modern representatives of these long-lived genera 
were considerably different from those of their Paleozoic namesakes, and students 
of the particular groups to which they belong would be loath to unite the fossils 
with their supposed modern representatives. In fact, only two of the ten Ordovician 
"immortals" have well-supported claims to their titles. One is the geologically famous 
inarticulate brachiopod, Lingula. Fossils have been found which show the mold 
of the fleshy pedicle, and others which retain imprints of all the numerous muscles 
8f the modern form. The other is the worm-tube, Spirorbis. Although there is less 
evidence in this case, the size, form, direction, and rate of coiling are so nearly identi- 
cal in Ordovician and recent specimens that there is no way of distinguishing between 
them. Both these animals appear to have survived because they could live under a 
variety of conditions. All that they needed was plenty of water. Both are found in 
marine, brackish water and in what seem to be fresh-water deposits. Attached to 
floating objects, they drifted far out to sea, but they were equally successful in with- 
standing the attacks of the surf upon rocky headlands or the shifting sands of 
beaches. Spirorbis is just as common on the late Paleozoic plants of the coal swamps 
as it is on the brown seaweeds of the present day. Lingula probably got into fresh 
water less often but still lives in estuaries under almost incredibly adverse conditions. 

The chief interest in the survival of such an animal as Lingula lies in the fact that 
it is one of the most primitive of all known brachiopods, and close to the central 
stock from which all the other members of that highly diversified phylum sprang. 
It has many descendants, some of them so different from itself that the connecting 
links have not yet been found. A second example of the survival of a primitive stock 
may be drawn from another family of brachiopods, an extraordinarily interesting 
group which has, for lack of space, been neglected in this book. The ones referred 


to are the spine-bearers, productids, most widely distributed and abundant animals 
of the second half of the Paleozoic. Their progenitor, Chonetes, was derived from 
an earlier spineless form in the late Ordovician. Chonctes has spines along the pos- 
terior margin of one of the two shells (valves) only. They appear to have served 
at first as organs of temporary fixation, but the sessile habit, thus initiated, was 

FIG. 154. Chart to show the times of culmination for certain groups, with a 
suggestion of their previous and subsequent history. 

quickly accepted. The Devonian Strophalosia, for instance, soon became fully spinose, 
and lost its freedom and the original symmetry of the shell. A somewhat more 
conservative race, the productellas of the Devonian, were attached to foreign objects 
by spines which grew from the larger (convex) valve only. They in turn seem to 
have been the ancestors of the cosmopolitan Productus, with spines on both valves. 
Although firmly anchored in the adult state, and "degenerate" in that they had lost 


the pedicle and various other structures of a typical articulate brachiopod, the produc- 
tids were the most successful members of their phylum. They produced the most 
numerous, the most varied, the most widespread, and the largest brachiopods of the 
late Paleozoic. It is probable that they gave rise to the Permian Richthofenia, an en- 
tirely sessile form attached by the apex of one shell only. Although Chonetes seems 
to have been the ancestor of all these spine-bearers, it survived as long as any of the 
descendants, longer indeed than some of them. Nor was its tribe one to be pitied 
as the feeble remnant of an ancient race. New species continued to spring from the 
central stock till the end of the Paleozoic, and individuals were just as numerous as 
the extraordinarily abundant Carboniferous productids, although less varied and of 
much smaller size. 

Many other instances of the same sort could be cited. They are important not 
only for their bearing on the future history of man but also as justification for our 
belief in "contemporaneous ancestors/* that is, in the survival of primitive types, 
which furnish information in cases where the geological record fails. 

The examples so far have been drawn from the invertebrates, for, since they 
change more slowly than vertebrates, they present clearer evidence. There are, how- 
ever, well-known instances among the higher animals. Attention has already been 
called to some of them. The cartilaginous ganoids gave rise to the bony ganoids, and 
they in turn to the teleosts, but the cartilaginous ones are still with us, although in 
reduced numbers. The turtles have changed but slightly since the Triassic; some of 
them have skulls not greatly different from those of the most primitive reptiles, the 
cotylosaurs. Old man turtle has kept paddling along, ignoring the changes in fashion 
which led to the destruction of dinosaurs, pterosaurs, ichthyosaurs, plesiosaurs, 
mosasaurs, and other groups. 

Few mammals have survived for lojig periods without considerable alteration. 
The opossum is a striking example, but it is the only surviving primitive marsupial. 
Some of the insectivores retain a primitive placental dentition, which harks back to 
Cretaceous times, but their limbs have been variously modified. Dogs represent the 
central line of the carnivores, although their modern representatives exceed the Eocene 
stock in complexity of brain and in size. Tapirs are often referred to as "living fos- 
sils." In some respects they are like the Eocene ancestors of various odd- and even-toed 
vegetarians. Nevertheless, although they retain some primitive characteristics, they 
can be called primitive only by comparison with such highly specialized relatives as 
the modern horses, rhinoceroses, and camels. 

Relict animals or plants are those which retain primitive characteristics despite 
the vicissitudes of the ages. First on the list are Lingula, Spirorbis, and the like, 
animals seemingly able to withstand all sorts of conditions within their own milieu, 
in this case the aquatic one. Turtles have done a little better, for they produced the 
terrestrial tortoises. Omnivorous mammals, such as the opossums and insectivores, 


belong in this group. Tapirs typify a second series, composed of organisms which 
have survived only because of an ignominious retreat to that part of the world where 
life is easiest, because food is most abundant. The arctic, sub-arctic, and cold-temperate 
zones of the present day contain no relicts of this type. Only vital groups can live 
there. It is in the warmer belt alone that such out-of-date animals as apes, monkeys, 
elephants, rhinoceroses, tapirs, lions, tigers, and the like survive. Man is not a relict, 
but an active, primitive creature. Like Lingula, he is able to live under all sorts of 

In this discussion, emphasis has been placed upon conservative tendencies in 
evolutionary processes. On the other hand, evolution has been described, as indicated 
by the real meaning of the word, as a process of unfolding. The fact that primitive 
creatures do survive shows that, if there be an inherent tendency toward differentia- 
tion, it can be inhibited, but the "unfolding" idea is to some extent supported by the 
facts of geological history. Each group shows a multiplicity of forms. There are, or 
have been, crawling reptiles, walking reptiles, flying reptiles, gliding reptiles, and 
swimming reptiles. Mammals show the same differentiation into forms which, exter- 
nally at least, closely resemble the reptiles occupying corresponding habitats. In fact, 
animals of all sorts have a tendency to occupy every zone and habit of life. This is 
what H. F. Osborn called the "Law of Adaptive Radiation." All animals and plants 
show it, the higher groups more fully than the lower. The sponges, for example, are 
mostly sessile; a few float, but none crawls of swims, and all have the same method of 
feeding. The coelenterates have been somewhat more successful, for, although domi- 
nantly sessile, some float, others swim, and at least two of them crawl. Fish have 
done still better, for they swim, float, crawl, jump, and almost fly. Their food habits 
are more diversified than those of other aquatic animals. Most are carnivorous, but 
many are herbivorous. The carnivores differ greatly in their diets. Some have 
piercing, some grasping, some cutting, and others crushing teeth. Terrestrial animals 
have greater opportunities than aquatic ones, but the earliest of them, the amphibians, 
show less diversification than the fishes. The first invertebrates to get onto the land, 
the diplopods, likewise failed to live up to their possibilities, but the success of their 
cousins, the insects, is well known. No greater radiation than theirs is imaginable; 
only at sea have they been dilatory. 

Radial evolution does seem to be a sort of unfolding, but not in the sense of a 
direct upward trend. In fact, much of the movement has been lateral or backward, 
so far as progress toward higher animals is concerned. Reptiles did not evolve to 
their maximum and then produce mammals (Fig. 155). Mammal-like reptiles were 
a side line, doomed to expire before the end of the Triassic; even they did not cul- 
minate when producing mammals. Their "great men" were big, foolish-looking 
creatures which managed to thrive in Africa, Europe, Asia, North and South America 
till the dinosaurs usurped their position at the end of the Triassic. Like the Israelites 
in their journey, the South American representatives of the group got across the Red 


Sea of North America without many casualties, whereas Pharaoh's pursuing army 
of dinosaurs was properly bogged down. Small, inconspicuous, and lowly were the 
members of the therapsids which gave rise to the mammals. We are not yet sure 
that we have identified them. 

The culmination of the reptiles was not in the mammals but in the snakes, the 
most specialized of all. As has been pointed out, the whales are the highest of the 
mammals, and the birds are at the summit of the whole animal kingdom not man. 

A vase is begun; why, as the wheel goes around, does it turn out a pitcher? 

Horace grasped, to a certain extent, the seeming casualness of evolution. If there has 
been any purposefulness, any inherent direction, it is obscured by radiative adapta- 
tion. If nature tried to produce in man the perfect vase, all she achieved was a pitcher. 

Most chroniclers of the history of life begin or end with a "family tree," with 
roots in the protozoans, a trunk of invertebrates, and a series of modern animals 
ornamenting the branches. I have only an elementary knowledge of botany, but 
I doubt if there ever was such a plant. A truer idea of the paths of evolutionary 
change may perhaps be gained from the accompanying diagram (Fig. 156), in 
which is expressed the idea that although on the whole there has been progress in 
an upward direction, it has not been straight forward but is the result of the response 
of animals to* all conditions of environment. In other words, progress has been the 
more or less fortuitous result of radiative adaptation. 

Men like to think that the evolution of the earth and its inhabitants was pur- 
poseful, which probably accounts for the popularity of the theory of orthogenesis, or 
"straight-line" evolution. In dealing with the histories of the horse and the camel 
we followed to the present day lines of descent which showed persistent, although 
not continuous, change in the direction of lengthening of the legs and feet, reduction 
in the number of toes, and increasing complexity of teeth. On the other hand, the 
central, direct line of the rhinoceroses perished in the Pliocene; it did not lead to the 
modern forms, which must have been derived from some lateral branch whose 
family records have not yet been discovered. And, as was intimated in the same 
chapter, there were various other lineages of horses and camels than the ones described. 
That is, evolution was radial within each of these groups. As we look still further 
back, we find that the stock from which the horse arose also gave rise, on the one 
hand, to the tapirs, the rhinoceroses, and the titanotheres, and, on the other, to various 
lines of even-toed mammals. Radiation was superimposed upon radiation. In other 
words, an orthogenetic line is one that is man-picked. Every living individual has 
an ancestry extending straight from himself to the first living protoplasmic particle. 
Each living creature, coral, butterfly, fern, king, or criminal, is the culmination of 
an orthogenetic series. Whether a man be a king or a criminal (there have been 
plenty of instances where the two were combined in one person) depends upon two 



things, heredity and environment. It is still a question which is the more important. 
The paleontologist, partly from training, partly from his knowledge of ancient history, 
is apt to stress the latter. 


- Infusorians o 

/A I 


- Anthozoans > 

r~i i 


Annelids z 
































f 1 
























































FIG. 155. A diagram showing the geological ranges of certain groups, ^ 
to emphasize the fact that each line has evolved in its own way, and not in 
the direction of the next more highly organized group. As indicated, higher t 
groups arise at the beginning, not late in the history of any phylum. The 
term Coelomate Invertebrates as used here refers to the primitive ones only. * 
The chart was not intended to show all the lines of invertebrates. 

The times of greatest change in plants and animals that is, of most rapid 
evolution have coincided with the great physical events in earth history, those which 
have drastically altered the distribution of land and sea, raised mountains, or partially 


submerged continents. Quiet intervals have in general been periods of relative stag- 
nation for organisms. Not that change ever absolutely ceases, for, no matter how 
uniform the conditions under which animals and plants live, the "struggle for ex- 
istence'* is perpetual. The three important factors of environment are geographic 
conditions, food, and ecological relationships. The most striking changes in organ- 
isms occur when there are marked alterations of the first, for the others are de- 
pendent upon them. But even though there be no earth movements, physical con- 
ditions are not always the same. There are periods of heat and cold, moisture and 
drought, changes in the salinity of the seas, in position of oceanic currents, et cetera. 
Even extraterrestrial forces, such as sun spots, have their influence. 

The associations of organisms with each other in connection with their habitat 
that is, their ecology must have been of great importance in evolution, but so little 
is known of paleoecology that it is impossible to evaluate that side of the subject. The 
overproduction of individuals by some groups is merely a bit of luck for others. 
Food is thereby provided. If the sea were populated entirely by oysters, and the only 
land plants were weeds, man might survive, but he would probably be a discontented 
mammal. Nature has somehow struck a balance favorable to the present proprietors 
of the earth. 

For animals, food is probably the most important of the three factors mentioned 
above. As far back as the Cambrian, food played a leading role, for it was evidently 
its presence on the sea floor that led some of the animals to adopt a sessile or motile 
benthonic existence. The absence of any great quantity of food on land greatly de- 
layed the evolution of terrestrial animals. As has been shown, the first tetrapods 
were not herbivores, but carnivores. 

Organisms appear to have reached the land from the sea by way of estuaries 
and rivers. The present paucity of the faunas of fresh waters proves that few groups 
have established themselves there. It is probable that the ostracoderms were among 
the first, and their jawless condition shows that they did not enter the new habitat 
in pursuit of any large or active prey. Microscopic plants and animals or decaying 
green algae probably satisfied their appetites. All except the anaspids appear to have 
been mud-grubbers. The earliest gnathostomes are the Lower Devonian acanthodians, 
and the Mid-Devonian ganoids and lung fish. Their oldest representatives are so 
highly organized that it seems probable that these groups originated during Silurian 
times. ^Perhaps the first real jaws were those which chewed graptolites and soft- 
shelled trilobites. The Paleozoic amphibians were carnivores, dependent chiefly upon 
such food as they could obtain from the waters of streams, lakes, and swamps. Their 
somewhat more terrestrial offspring, the primitive reptiles, had much the same 
habits, although there were among them a few which seem to have been satisfied 
with a diet of insects, insect larvae, worms, and other inhabitants of the land. Appar- 
ently the higher animals were reluctant to try a vegetable diet until necessity forced 


it upon them in Mesozoic and Tertiary times. So long as vertebrates lived in the 
water, the scheme seems to have been to expect the more lowly invertebrates to feed 
upon simple aquatic plants, for fish to eat the invertebrates, and for the lordly ter- 
restrial vertebrates to devour the fish, or each other. Until recent times "Devil take 
the hindermost" was the general custom. No wonder that this rule dominates so 
large a* proportion of human beings. 

FIG. 156. A diagram showing the probable relationships of the important 
groups of animals, in the progress upward from the Protozoa. The radial 
lines about the names indicate that in each group there has been evolution 
in all possible directions, but only a few lines have led to progress toward a 
higher group. Coelomata should be read as "Primitive Coelomata.*' This an 
attempt to replace the familiar but misleading "family tree/' 

The beginnings of vegetarianism among the higher animals must have long ante- 
dated the time at which true vegetarians appear in the geological record. It seems 
probable that the sequence was in each case from carnivore through omnivore to 
herbivore. This was the history of the herbivorous mammals, and it was probably 
that of the other groups. As has already been noted, certain Pennsylvanian gymno- 
sperms, Cordaites, produced tempting fruit, and Triassic and Jurassic members of 
the same group bore seeds and fruits. These inducements led various animals to 
essay arboreal life, which culminated in the evolution of the birds, pterosaurs, early 


arboreal insectivores, and bats. But the climbers carried with them into the trees a 
taste for flesh, and their new habitat provided abundant food of this kind, larval 
and adult insects, eggs and youthful progeny of birds, and various other odds and 
ends. Hence most arboreal creatures are naturally more or less omnivorous. In 
general, arboreal life and its resultant diet has led to loss of teeth. For some reason 
this has affected mammals less than birds or reptiles. Fortunately man got down, 
or was driven down, from the trees before he became a monkey "driven down," 
not by climatic changes, but because he was a more primitive, less competent, and 
less pugnacious animal than his associates. (This is an idea new to the writer, having 
occurred to him during the few seconds since he began to write the previous 'sentence. 
Perhaps man owes his present position to his inferiority to the other anthropoids: 
without large canines he could not fight, and his short arms and fingers made him an 
unsuccessful brachiator. I shall not try to develop the theme, but merely list it as a 
possibility. As I frequently tell my students, the only way to orient oneself is by 
writing. Merely guide the nib, and ideas flow with the ink not that the ideas are 
all good.) 

The mammals responded more rapidly to changes in the supply of food than any 
other vertebrates; in this group alone are there numerous herbivores. Among modern 
reptiles all, except for the omnivorous turtles, are carnivores. Among extinct ones 
the sauropods and the herbivorous dinosaurs alone essayed a different food. The 
former were probably omnivorous; the latter never got beyond the browsing stage. 
This was not entirely their fault, for they expired just as grasses came into the fields. 
It is well said that it is difficult to teach an old dog new tricks. The ceratopsians may 
have died in the midst of plenty, just as a horse would if turned loose in a coniferous 

In contrast to the earliest reptiles, which were carnivorous, those early mammals 
which left descendants were omnivorous. They merely followed the paths of least 
resistance when they accepted whatever food nature offered them. It can hardly be 
said that certain mammals were predestined to certain diets, and that they followed 
routes of evolution already picked out for them. The evidence seems to be clear 
that the habits of the animals changed with the ever-varying nature of the food. The 
placental mammals were a new and plastic line in late Cretaceous and early Tertiary 
times. Their habits had not been fixed, as were those of the reptiles at the dawn 
of the JVlesozoic. The group was still in its infancy, still capable of variation and 
adaptation. Man has inherited this primitive condition; all is grist which comes to 
his mill. He thrives, despite fads or fancies. 

The geological record seems to show that the progress of evolution has been 
governed largely by the physical environment. From the beginning, life has de- 
pended on external circumstances and more or less accidental coincidences. As Law- 
rence Henderson has shown in his book, The Fitness of the Environment, life as we 


know it could not exist but for the fact that the most abundant liquid on the earth is 
water, the dominant gas, oxygen, diluted in proper proportion by inert nitrogen, and 
that carbon dioxide is present. 

Nothing is really known about the origin of life, but it seems probable that its 
inception was the result of a huge chemical experiment, during which conditions 
were, such as had never previously obtained on this earth and will never be repeated. 
Whether one believes in a hot nebular or a cold planetesimal origin for the earth, 
there must have been a period when the outer zone was molten. During the period 
of cooling, chemical combinations may have been brought about which cannot be 
duplicated on the small scale of the modern experimental laboratory. Once life 
came into existence, it might have gone on forever in lowly forms had it not been 
for the constant driving of external forces. 

Evolutionary change seems to have been slow while animals remained in the 
water, a more or less static environment. As has been pointed out, it appears to have 
been speeded up when organisms reached the land, and to have proceeded most 
rapidly during the Tertiary. Nevertheless, there has been no definite rate of accelera- 
tion. Loomis tells us that the change from Eohippus to Orohippus, involving only 
the addition of a cusp on the third premolar, is spread over a period of two or three 
million years. The loss of the fourth digit of the front foot of the horses was accom- 
plished in about ten million years. Yet Merychippus arose from Parahippus, and in 
its turn gave rise to Hipparion, Pliohippus, and Protohippus withiii a brief period in 
the early Miocene. The rate varies with external circumstances. 

One of the most striking features of evolution from the end of the Mesozoic onr 
ward has been the general increase in the relative size of the brain of all vertebrates 
which have survived. Education appears here, either as cause or effect; it is probable 
that it was the former. 

During the Paleozoic, when fertilization was largely external, the mother laid 
her eggs in the water or on the sand, and, with that, responsibility ended. Whether 
impregnation occurred before or after the extrusion of the eggs, the father was not 
interested in the outcome. In the course of time, however, some animals began to 
protect their eggs, and there was gradually evolved the habit of retaining them within 
the body of the mother until the time of hatching. Thus arose what is called the 
mother instinct. Just when this began is not known. Probably not before Mesozoic 
times. (Only one of the many trillions of trillions of eggs laid during the Paleozoic 
is now known. It occupies an inconspicuous position in the Museum of Comparative 
Zoology at Harvard. Perhaps it is the only one which failed to hatch. It had lain 
buried in the muds of the Texan Permian till Llewellyn Price routed it out.) The 
mothers among the higher reptiles, birds, and mammals feed and train their children 
during the early days, weeks, or months of their lives, giving them a decided advantage 
over the hosts of more primitive creatures left by their parents to learn by trial and 


error, if at all. As Professor Hervey W. Shimer has pointed out, such education led 
to the keeping together of the family and so to the formation of societies, with social 
(communal) behavior. 

What does the history of life teach? Simply this: that man is a part of nature, 
physically governed by forces which are as yet only partially within his control. Real 
progress will be made only when facts are faced. Man is an animal, and a badly 
assembled one at that. Nevertheless he differs from others in that he is articulate 
and can be made literate. He alone is conscious of his position in the world, and 
that largely because of written records. Cats and dogs and birds and innumerable 
other animals teach their young enough to enable them to get along more or less 
successfully, but they lack foresight because they have but a limited hindsight. Man 
will, like the dinosaurs, follow a royal road to destruction unless he heeds the warn- 
ings of past history. Just at present he appears to be heading toward another dark 
period, like that of the Middle Ages. Despotism seems to be coming back into 
fashion. Let us hope it is nothing more than a threat, for with it would come all the 
evils of the feudal system. Like any other animal, man prefers to be ruled, bossed, 
and herded, because it saves him the trouble of thinking. Thousands of years may 
pass before he learns to use his single unique organ, the specialized brain. But I 
do not wish to be unduly pessimistic. There is still abundant time to change. At the 
worst, we know that dark ages have come and gone. Men have built new structures 
on the ruins of the past. Records of some sort survive, and much is learned from 
them. There will come a time when there will be only one language, when narrow 
patriotisms born of isolation will be forgotten, and when increase in population will 
be adjusted to food supply. Will that be the millennium? Maybe, but only if man 
lives up to his possibilities. I'll never know, nor will you. Let us hope that man will 
learn to direct his future by his knowledge of the past. 



Abbott, G., 31 

Abel, Othcnio, 138, 169, 191 

Acanthodes, 92 

Acanthodians, 89, 91-93, 106, Figs. 49, 52 

Aceratheres, 257, Fig. 135 

Adapts, 277 

Adelospondyli, 112, 115 

Aepyornis, 183, Fig. 92 

Aglaspis, 55, 56, 58 

Aigialosaurs, 159 

Air-breathing, initiated, in the arthropods, 60, 

103; in the vertebrates, 99, 103 
Aistopoda, 114, Fig. 63 
Alder flies, 204 
Algae, 25; blue-green, 29, 30, 32, 293; brown, 

32, 293; green, 293; red, 25, 32, 293 
Alligators, 125, 129, 144, 164 
Allosaurus, 132, 141 
Allotheria, described, 213, 214, 222, 223, Fig. 


Amblypods, 223, 225, 226, Fig. 116 
Amebelodon, Figs. 141, 144 
Ammonoids, described, 194-199, Figs. 98-101; 

range, 302, Fig. 154 
Ammonites, see Ammonoids 
Amoeba, 5 
Amphibians, 22; characteristics, 108, 109; 

crossopterygian characteristics of, 103, 104; 

described, 108-118; footprints, 109; meta- 
morphosis, 108; oldest, 109; origin, 97-106; 

range, 301 
Amphinome, Fig. 9 
Amphioxus, 84, 85, 95; theory of origin of 

vertebrates, 84 
Amphitherium, 216 
Anapsids, 125, Fig. 66A 
Anaspids, described, 80, 95, Figs. 45, 49; theory 

of origin of vertebrates, 88-93, 95 
Anchisaurus, 131 
Ancyloceracones, 198, Fig. 100 
Andrews, Roy Chapman, 142, 217 
Andre wsarchus, 229 
Andrias scheuchzeri, 116 
Angiosperms, 298, 299 
Animals and plants, differences between, 4 

Annelids, 32, Figs. 7, 9; larva, 86, Fig. 48A; 

theory of origin of vertebrates, 86, 95 
Annularia, 296 

Annulata, 20, 24, 28, Figs. 7, 9 
Anteaters, 266 
Antelopes, 251, 252 
Anthropoids, 276, 278 
Antiarcha, 87, 88, 106; described, 80, 81, 

Fig- 44 

Antilocapridae, 252 
Ants, 208 

Anura, 108, 116, 118 
Apterygota, 202 
Apteryx, 183 

Arachnids, 22; described, 55-61 
Archaeocyathinids, 24, 36, Fig. 8 
Archaeopteryx, 173, 176, 183, 184, 188, Figs. 

9 r 93> 95 
Archaeornis, 174, 176, 183, 184, 186-188, 

Fig. 88 

Archelon, 160, 161, Fig. 84 
Archidis\odon, Fig. 141 
Arctodus, 232 
Arctotherium, 232 

Armadillos, 226; described, 266-268 
Arthrodires, 88; described, 81-83, Figs. 46, 47 
Arthropods, 21, 23, 27, 28, 33, 47, 55; theories 

of origin of vertebrates, 87, 88 
Artiodactyls, 226, 239; described, 243-252, 

Figs. i23A,D; evolutionary trends of, 243, 


Aster oxylon, Fig. 150 
Astrapotheria, 223 
Atifofywia, 31 
Auchenaspis, Fig. 45 
Australopithecus, 287 
Aves, see Birds 

Axonolipa, 43, 45, Figs. 15, 17, I9C,D,E 
Axonophora, 43, 45, 46, Figs. 18, I9F,G 

Bacteria, 25, 30; importance, 292, 293 

Bactriticones, 198, Fig. 100 

Baculiticones, 198, Fig. 100 

Balanoglossus, 84-86, Fig. 48; larva, 85, 86, 

94, Fig. 4 8B 
Baluchitherium, 243, 258, Fig. 137 



Bark lice, 205 

Barnacles, 21, 40 

Barrande, Joachim, 22, 53 

Barrell, Joseph, 101-103, 291 

Bar y lambda, Fig. 116 

Bather, Francis A., 72 

Bats, 223, 224 

Bears, 229, 231, 232, Fig. 119 

Beebe, William, 184 

Beecher, Charles E., 123, 127 

Bees, 208 

Beetles, 205, 207 

Belemnites, 190, 191, Figs. 96, 97 

Beltina danai, 31, 35 

Birds, 22; ancestors, 183-188; characteristics, 

174, 175; rare as fossils, 174; reptilian char- 

acteristics of, 173, 177; toothed, 14, 173-180; 

flight, origin of: cursorial theory, 183, 184; 

gliding theory, 185; tctrapteryx theory, 184, 

i8 5 

Bison, 265 

Blastoids, 20, 40, 62, 70, Fig. 37; range, 301, 
Fig. 154 

Blastomeryx, 250 

Blattaria, 202 

Bothriolepis, 80, 81, 87, 88, Fig. 44 

Bovidae, 252 

Brachiopods, 21-23, 27, 28, 33, 36, 42; articu- 
late, 21, Fig. I2A,B; inarticulate, 21, Fig. u 

Brachiosaurs, 136 

Brain, regions of, 75 

Branchiosaurs, no, 112, 113, Figs. 57, 59, 61, 
62; metamorphosis, 113 

Brittle $#rs, 20, 68, Fig. 36 

Bronto/aurus, 18, 132, 148; described, 135, 136 

Brooks, W. K., 34, 35 

Broom, Robert, 98, 102, 103, 145, 187, 211, 212, 

Brown, Barnum, 138 

Brues, C. T., 208 

Bryant, W. L., 77, 98, 105 

Bryozoans, 20, 22, 23, 27, 40, Figs. 9, 12 

Bugs, 207. 

Bulman, O. M. B., 42 

Bunodonts, described, 244-246 

Butterflies, 207 

Cacops, 112, Fig. 60 
Caddis flies, 207 
Catamites, 296 

Camels, 227; evolution, 258-263; foot, Fig. 
1 23 A 

Campbell, Douglas, 295 

Camptosaurus, 137, Fig. 77 

Cants familiaris, 230, 231 

Cams lupus, 230, 231 

Carnivora, 223; described, 228-237 

Carpenter, Frank M., 200, 202, 208 

Case, E. C., 127 

Casts, 6, Figs, i, 3 

Catarr nines, 276, 278 

Cats, 229, 230, 233-236, Figs. 120, 121 

Caudata, 108 

Cavicornia, 249; described, 251, 252 

Cayeux, L., 30, 31 

Cephalaspids, 78, 79, 81, 88, 94, 106, Figs. 43, 


Cephalaspis, 79, Fig. 43 
Cephalopods, 21, 37, 40; described, 189-199, 

Figs. 96-101; phytogeny, 198, Fig. 101; 

shell forms, Fig. 100 
Ceratopsia, 130, 146, 150, 162; described, 


Cervicornia, described, 249, 250, Fig. 130 
Chaetopoda, 24, 33, Figs. 7, 9 
Chalicotheres, described, 242, 243, Fig. I23C; 

exceptions to Cuvier's law, 242 
Cheirocrinus, Fig. 35 
Chilopoda, 22 

Chimpanzee, 278, 283, 286-288, 290 
Chitons, 40 

Choeromorus, Fig. 126 
Chonetes, 303, 304 

Chorda ta, 11, 22, 23, 25, 27, 41, 73, 84, 86 
Civets, 229 
Clark, A. H., 66 
Clark, H. L., 66 
Clathrotitan, Fig. 105 
Cleland, H. F., 45 
Climatius, 92 
Cobb, Irvin S., 39 
Coccosteus, 81, 82, Fig. 46 
Cockerell, T. D. A., 208 
Cockroaches, 202, 203 
Coconuts, Fig. 153 
Codfish, Fig. 50 

Coelenterates, 20, 23, 24, 26-28, 32, 33, 40, 62 
Coelomates, 26, 27, Figs. I3E,F 
Coelurosaurs, 134, 145, 146, 150, 186 
Coleoptcra, 204, 205 
Collembola, 202, 204, Fig. 103 
Compsognathus, 133, 186 
Comstock, J. H., 201 
Condylarths, 223, 225, Fig. 114 



Cook, Harold, 258 

Cope, Edward Drinker, 17, 120, 124, 127, 302 

Corals, 20, 33, 40, Fig. I2D; described, 62; 
extinction of, 65; formation of skeleton, 62, 
63; history, 65; larva, Fig. 326; symmetry, 

Cordaitcs, 297, 309 

Corythosautus, 139 

Cotylosaurs, 121-124, I2 ^> Fig s - 64, 65 

Credncr, H., 113 

Creodonts, 225, 229, 230 

Crinoids, 20, 40, 62; camcratc, 69, Fig. 34; 
described, 68; free-swimming, Fig. 33; his- 
tory, 69; range, 301, Fig. 194; stalked, Fig. 


Crioceracones, 198, Fig. 100 
Crocodiles, 125, 129, 144, 145, 149, 162; marine, 


Cro-Magnons, 284, 285 
Crossopterygians, 97, 102, 106, Figs. 53-56; 

amphibian characteristics of, 103, 104 
Crustaceans, 22, 23, 32, 37, 55, Fig. 10 
CryptolithuSy 51-53, Fig. 21 
Cryptozoon, 30, 32 
Cuttlefish, 189 

Cuvier, G. L. C. F. D., 128, 168, 242, 269 
Cynodictis, 231, Fig. 118 
Cynodonts, 212, 214, 215, 219 
Cynognathus, 211, 214, 219 
Cyon, 271 

Cyrtoceracones, 194, Fig. 100 
Cystids, 20, 23, 24, 33, 36, 62; as ancestors, 71; 

described, 70, 71, Fig. 35; range, 301, Fig. 

Cystoids, sec Cystids 

Daly, Reginald A., 33, 34, 36, 65 
Daphaenus, see Daphoenus 
Daphocnus, 232-233 
Darrah, William C., 294 
Dart, Raymond, 289 
Darwin, Charles, 253 
David, Sir Edgcworth, 32 
Dawson, William, 282 
Dcane, James, 128, 129 
De Chardin, Father Tcilhard, 282 
Deer, 227; history, 249, 250 
Deinosauria, see Dinosaurs 
Dendroidea, 43-45, Figs. 17, i9A,B 
Dermaptera, 207 
Dermoptera, 223, 225 
Diacodexis, 261 

Diacodon, Fig. 115 

Diapsids, 125, 126, Fig. 

Diatoms, 25; as food, 293, 294 

Diatryma, 180, 181, Fig. 90 

Dibranchiata, 189, 190, Fig. 96 

Dicer athcres, 257, Fig. 136 

Dicer atops, 142 

Dicotyledons, 298 

Dictyoncma, Fig. 17 

Dimetrodon, 127 

Dimorphodon y 171 

Dinichthys, 81-83, 106, Fig. 47 

Dinictis, 234, 236, Fig. 120 A 

Dinohyus, 245, Fig. 127 

Dinar nis maximus, 182 

Dinosaurs, 125; ancestry, 145; bones in quarry, 
Fig. 5; brains, 148, 149; causes of extinction, 
148-152, 298, 299; classification, 130; de- 
scribed, 128-152; food, 150; origin of bi- 
pedality, 146-148; possible ancestors of birds, 
185, 1 86; range, 302, Fig. 154; sacral brain, 
149; tracks, 128, 129, Fig. i; where found, 

DiplodocuSy 18; described, 135, 136, 148, Fig. 


Diplopoda, 22, 103 
Diplovertebron, 123, Figs. 57, 58 
Dipnoi, 97, 100 
Diprotodonta, 215 
Diptera, 201, 207 

Dogs, 229-231; genealogy, Fig. 119 
Dolichosoma, 115, Fig. 63 
Dollo, L., 182 
Draco, 164 
Dragonflies, gigantic, 203; true, 204, 207, Fig. 


Drepanaspis, 76-78, Fig. 42 
Dromatherium, 214, Fig. 109 
Dromomeryx, 250, Fig. 130 
Dryopithecus, 278, 287-290, Figs. 1476, 1486 
Dubois, Eugene, 279 

Earwigs, 207 

Eaton, George, 168 

Echinoderms, 20, 23, 27, 28, 32, 40, 62; as 

ancestors of vertebrates, 93, 94; described, 


Edaphosaurus, 127, Fig. 70 
Edentates, 223, 226; described, 266-269, Fig. 


Edrioasteroids, 20, 23, 24, 33, 71, Fig. 36 
Edwardsia, 66, Fig. 32A 



Elasmobranchs, 74, 78, Fig. 39 

Elephants, 227; history, 264, 265, Fig. 141 

ElephaSy 279, Fig. 141 

Elginia, 123, Fig. 6$A. 

Embiaria, 204 

Embolomeri, in, 123, Figs. 57, 58, 59A,B 

EmpcroccraSy 198 

Emplectopterisy Fig. 152 

Entclodonts, described, 245, 246, Fig. 127 

Eoanthropus dawsoni, 282, 283, 287-289, Fig. 


Eodelphisy 217, Fig. in 
EogyrinuSy 105 

EohippuSy 254-257, 311, Figs. 131, 140 
Eoliths, 282 

EosauravuSy 120, Fig. 69 
Eozoon, 32 
EpihippuSy 254 
Eporeodon, 246 
EquuSy 254, 255, 266, 271, Fig. 131; E. caballus, 

266; E. giganteuSy 266 
EryopSy in, Fig. 64A 
Eudendrium, Fig. 16 
Euparfyria, 145, 187, Figs. 93, 94 
Eurypterids, 32, 40, 87, 88; described, 57, 58; 

habitat, 58, 59 
Eurypterus, 31, 57, Fig. 28 
Eusarcus, 58, 60, Fig. 30 
Eusthenopteroriy 98, 102, 104, Figs. 53, 54, 56 
EuthacanthuSy 92 
Evolutionary change, times of, 307, 308; rate 

of, 25-28, 311 

Favosites, 64, 66, Fig. 38 

Felidae, sec Cats 

Felts, 10, 234, Fig. I20C; F. tf/ro*, 235; 

F. catusy 10; F. A?o, 10; F. tigrisy 10 
Ferns, 296, Fig. 4 
Fish, 14; bony, 74, Figs. 39~4i; fins, 75; 

scales, 74; skeleton, 73, 74 
Fissipcdia, 229, 230 
Flies, 201, 207 

Food, as a factor in evolution, 308-310 
Foraminifera, 19, 24, 30, 33, Fig. 6 
Fossils, collecting, 12-18; definition, 8; names, 

9-11; processes of preservation, 5; states of 

preservation, 7; study, 8; submarine, 13; 

where found, 12 

Ganoids, 74, 75, Fig. 39 
Gaskcll, W. H., 87 
Gastroliths, 149 

Gastropods, 21, 23, 32, 33, 36, 40; range, 301, 

Fig. 154 

Geological timetable, xi 
Gibbons, 278 

Gilmore, Charles W., 136, 137 
Giraffes, 249, 250 
Giyptodonts, 266, 268, Fig. 143 
Goats, 252 

Goniatites, 196, Fig. 98 
GorgosauruSy 134, Fig. 76 
Gorillas, 278, 286, 288, 290 
Grabau, A. W., 58 
Granger, Walter, 142 
Graptolites, 20, 33, 40; described, 42, Figs. 15, 

17-19; growth of colonies, 43, 44; habitat, 

44; history, 45, 46 
GraptolithuSy 42 
Grasshoppers, 203 
Gregory, William K., 142, 184, 185, 215, 


Gymnosperms, 297, 298 
Gyroceracones, 194, Fig. 100 

Hallo pus, 132 

Haly sites y 64 

Handlirsch, Anton, 200 

Harrimania, 85 

Harmer, S. F., 85 

Hay, O. P., 269 

Hedgehogs, 221, 223 

Heilmann, Gerhard, 177, 185-188 

Heintz, Anton, 82 

Hemichordata, 85 

Hemicyoriy 232 

Henderson, Lawrence, 310 

Heptodony Fig. I25A 

Hesperornis, 177-180, 182, Fig. 89 

Heteroptera, 207 

Hexacorals, 63-66 

Hickory nuts, Fig. 153 

Hippariony 311 

Hippopotamus y 244, 279 

Hitchcock, Edward, 128, 129 

Holm, Gerhard, 42, 59 

Holtedahl, Olaf, 30, 31 

Hominidae, 11, 276, 286 

Homo heidelbergensisy 280, Fig. I47C 

Homo ncanderthalensiSy see Neanderthal man 

Homo rhodesiensisy 289 

Homo sapienSy 11, 208, 256, 275, 283, 288, 300, 

Figs. i48D,G 
Homogalax, 257 



Homoptera, 204, 205 
Hophphoneus, 236, Fig. 121 A 
Horse, evolution, 253-256 
Horseshoe crab, 55, 56 
Huntington, Ellsworth, 147 
Huxley, Thomas Henry, 274 
Hyaenarctos, 232 
Hydrct, 42 

Hydroids, 20, 42, Fig. 16 
Hydrozoans, 33, 37 
Hyenas, 229 
Hymenoptera, 207, 208 
Hyolithes, Fig. 10 
Hyolithids, 33 
Hypsilophodon, 138 
Hyrachyus, 256, 257, Fig. 1236 
Hyracodon y 257 
Hyracodonts, 256 
Hyracotherium, 256 
Hysterogenicones, 198, Fig. 100 

Ichnology, 129 

Ichthyornis, 177-179, 182, Fig. 92 

Ichthyosaurs, 125, 160; described, 153-155, 

Figs. 79, 81 
Ichthyostega, 109 
Ichthyostegopsis, 109, Fig. 556 
Ictidosaurus, 212 
Iguanodon, 137, Fig. 73 
Implements of primitive man, 282 
Insecti vores, 215, 216, 218-220, 222, 229, 304, 

Figs. 112, 115; ancestors probably arboreal, 

Insects, 41; characteristics, 200, 201; described, 

200-208; effects of Permian glaciation on, 
204, 207; in amber, 208, Fig. i; oldest, 202; 
origin of metamorphosis, 207; origin of 
wings, 205; venation of wings, 201, 202, 
Fig. 104 
Isograptus, Fig. 15 

Jackson, Robert T., 126 
Jefferson, Thomas, 268 
Jellyfish, 20, 23, 24, 33, 36 

Keith, Sir Arthur, 285, 290 
Kiaer, Johan, 80, 91 
Klaatsch, H., 286, 290 
Knowlton, T. H., 147 
Koch, Lauge, 109 

Labyrinthodonts, 117, 121; described, 110-112 

Lacewings, 204 

Lakes, Arthur, 16 

Lambe, Lawrence M., 142 

Lamcere, A., 201 

LanarJ(ta 9 77, 78 

Lane, Alfred C., 33, 35 

Lankester, Sir Ray, 60, 286 

Lapworth, Charles, 44 

Lasanius, 89-93, Fig. 51 

Leidy, Joseph, 248 

Lemmatophorcty 206, Fig. 104 

Lemuroidea, see Lemurs 

Lemurs, 224, 288; described, 276, 277, Fig. 


Lepidodcndron, 296, Figs. 4, 151 
Lepidoptera, 207 
Lepidosiren, 99 
Lepospondyli, 113 
Umulus, 55-57, Fig. 26 
Lingula, 21, 302, 304, 305, Fig. 11 
Linnaeus, Carolus (Carl von Linne), 9, n, 


Uthothamnion, 293 
Litopterna, 223 
Lituiticones, 194, Fig. 100 
Lizards, 159, 162, 164 
Llamas, 262 
Lobe fins, 97, 101, 106 
Logan, Sir William, 29 
Loomis, Frederick B*, 231, 244, 249, 250 
Lophiodon, Fig. 124 
Loxodonta, 265 
Lucas, F. A., 183 
Lucas, O., 17 

Lull, Richard Swann, 147, 149 
Lungfish, 97, 102, 106; habits of, 99 
Lysorophus, 115 

MacBride, E. W., 37 

Machairodontidae, see Saber-toothed tigers 

Mac hair odus, 235, Fig. 1218 

Mac^enzia 9 66, Fig. 32C 

Mammals, 22; archaic, 209-219; characteris- 
tics, 209-211; Cretaceous, 217-219; fossil, 
where found, 15; Jurassic, 214-217; Meso- 
zoic, 213-219; Northern, invade South 
America, 271; origin, 211, 212; Paleocene, 
221-227; Pleistocene, 264-273; Southern, 
invade North America, 266, 268, 269; sub- 
classes, 210, 21 1 ; teeth, 210 

Mammonteus, 265, Fig. 141 

Mammuthus, Fig. 141 



Man, ancestry, 274-291; characteristics of 
skeleton, 274-276; driven from the trees, 
310; "expectation of life," 300-306 

Mann, Albert, 30 

Mantell, G. A., 128 

Marsh, O. C., 16, 174, 253 

Marsupials, 210, 215, 216, 218, 219, 222, 223 

Martens, 229 

Martynov, A. V., 200 

Mastodons, 264, 265, 271, Fig. 141 

Matthew, William Dillcr, 232, 236, 253, 267 

Mayflies, 205 

Mecoptera, 204, 207 

Megalobatrachus, 116 

Megalonyx, 269 

Megaloptera, 204 

Megalosaurs, 145, 146 

Meganura, 203 

Megasecoptera, 203, 205 

Megatherium, 269 

Merriam, f. C., 156, 236 

Merychippus, 255, 311, Fig. 131 

Merychyus, 247, Fig. 129 

Merycodus, 250 

Merycoidodon, 246, Fig. 128 

Merycoidodonts, 227; carnivore-like character- 
istics, 248; described, 246-248 

Mesohippus, 254, 255, 257, Figs. 131, 132 

Mesosaurus, 156, Fig. 80 

Metazoa, 25 

Miacidae, 229, 231, 233 

Miacis, 231, 233, Fig. 119 

Micrococcus, 30 

Microsaurs, 112-115, 117 

Mimoceracones, 198, Fig. 100 

MiohippuSy 255 

Miomastodon, Fig. 141 

Miotapirus, Figs. 125, I2<^E 

Mixosaurs, 156 

Moas, 182, 183 

Mocntherium, 264, Fig. 141 

Molds, 6, Fig. 3 

Moles, 22 1,. 223 

Mollusca, 21, 23, 27, 28, 33, 40 

Monitors, 159 

Monoclonius, 142 

Monocotyledons, 298 

Monotr ernes, 210, 218 

Moodie, Roy L., 151 

Moody, Pliny, 128 

Moose, 250 

M or opus, 243, Fig. I23C 

Morosaurus, 149 

Morse, Edward S., 21 

Mosasaurs, 160, 162; described, 158, 159, 

Fig. 83 
Moths, 207 
Multitubcrculates, 213, 214, 218, 219, 222, 

Fig. 108 
Mylodon, 269, Fig. 143 

Nanosaurus, 137, 145 

Nautilicones, 194, Fig. 100 

Nautilus, 189-191, 194; described, 192, 193 

Neanderthal man, 280, 281, 284, 287-289, 291, 

Fig. I48F 

Needham, J. G., 201 
Neoceratodus, 99 
Neolenus, Fig. 21 
Neolitnulus, 56 
Neoliths, 282 
Neuroptera, 204, 207 
Nimravus, 234, Fig. 1206 
Nipponites, 198 

Nopcsa, Baron F., 150, 183, 184 
Notharctus, 276, Fig. I45A 
Nothocyon, 231 
Nothosaurs, 158 
Notkrotherium, Fig. 143 
Notoungulata, 223 

O'Connell, Marjorie, 58 

Octopus, 189, 190 

Odonata, 204 

Okapi, 250 

Old age, racial, 161 

Ophiacodon, 126, Fig. 68 

Ophiocones, 198 

Ophiurans, 68, Fig. 36 

Opossums, 215, 217, 219, Fig. in 

Orangoutangs, 288 

Oreodons, see Merycoidodonts 

Ornithischia, 130, 136, 146, Fig. 71 

Ornitholestes, 134, 186 

Ornithomimus, 150 

Ornithopoda, 130, 150; described, 136-139 

Ornithosuchus, 145 

Orohippus, 254, 311 

Orthoceracones, 193, 194, Fig. 100 

Orthogenesis, 300 

Orthoptera, 207 

Osborn, Henry Fairfield, 30, 124, 305 

Osteolepis, 98, 100, 104, 105 

Ostracods, 40 



Ostracoderms, 88, 94; described, 76-82 

Ostrich, pelvis, Fig. 93 

Owen, Sir Richard, 128, 168, 182, 256 

Oxyclaenidae, 229 

OxydactyluSy 262 

Palaeanodon, 226 

Palaeodictyoptera, 202, 203, 205, Fig. 102 

Palaeoloxodon antiquus y 265 

PalaeomeryXy 249, 250 

Palaeoscincus y 141 

Palaeosimia, 287 

Palapteryxy 182 

PaleanthropiiSy 280, 287, 288 

PaleaspiSy 76 

PaleolimuluSy Fig. 26 

Paleoliths, 282 

Paleomastodon, 264, Fig. 141 

PaleopithecuSy 287 

Pantotheres, 216, 218, 219, 223, 228 

ParadoxideSy 48, Fig. 20 

ParahippuSy 255, 311, Fig. 133 

ParapithecuSy 279, 286-288, 290, Fig. 147 A 

Parapsids, 125, 126, Figs. 67A,B 

Parasuchians, 144 

PareiasauruSy 123 

ParelephaSy Figs. 141, 142 

Patten, William, 81, 87, 88 

Peccaries, 227, 266; described, 244, 245 

Pecora, described, 248-252 

Pelecypods, 21, 40; range, 301, Fig. 154 

Pelion, Fig. 59 

Pelycosaurs, 126, 127, 211, Fig. 68 

PentremiteSy 70, Fig. 37 

Perissodactyls, 226; characteristics, 239, 240, 

Figs. 1 236,0 ; described, 239-243, 253-258 
Perlaria, 204, 205 
Peterson, Olaf A., 247 
PhareoduSy Fig. 41 
PhenacoduSy Fig. 114 
Phiomicty 264, Fig. 141 
PhororhacoSy 181 
Phrynosoma, 123, Fig. 656 
Phyllospondyli, 112 
Pigs, 244, 245, Fig. 126 
PinacoceraSy 196, Fig. 98 
Pinnipedia, 229 
Pisces, 22 

Pithecanthropus erectus, 279, 280, 287, 288 
Placentals, 211, 215-216; dentition, 211; oldest 

known, 218, Fig. 112; place of origin, 222, 


Planipennia, 204 

Plants, as food, 220, 238, 293, 294, 297-299; 
Cambrian, 24, 25; emerge from the water, 
295; first fruit-bearing, 297; history, 292- 
299; late Paleozoic, 296, 297; Mid-Devonian, 
296; oldest terrestrial, 294; Prc-Cambrian, 

29, 30, 32 

PlateosauruSy 131, 134, 145, 146, Fig. 75 
Platyrrhines, 276 
Plectoptera, 204, 205 
Pleistocene glaciation, influence, 270, 271 
PlesippuSy 255 
Plesiosaurs, 125, 160; described, 156, 157, Figs. 


Pliatichenidy 262 
PliohippuSy 255, 311 
PliopithecuSy 278, 287 
Pocock, R. J., 60 

Podofesaurus holyokensis, 131, 132, 134, 186 
Poebrothcrium, 261, Fig. 138 
Polyprotodonta, 215 
Porcupine, 223, 266 
Porifera, sec Sponges 
PortheuSy Fig. 40 
Predentata, 130, 145, 146 
Price, Llewellyn, 311 

Primates, 223, 224, 275; subdivisions, 276 
Primordial fauna, 22 
ProavtSy 188 

ProcameluSy 262, Fig. 139 
ProcompsognathuSy 132, 186 
Productids, 303, 304 
PromerycochoeruSy 247 
Prongbuck, 251 

Propliopithecusy 278, 286-288, 290, Fig. I48A 
ProtapiruSy 241, Figs. I25A,C,D 
Protelytroriy Fig. 102 
Protentomobrydy Fig. 103 
ProteosauruSy 153 
ProtoceratopSy 142, 146, Fig. 78 
Protodonota, 203-205 
Proelytroptera, 205, Fig. 102 
Protohcmiptera, 202, Fig. 105 
ProtohippuSy 311 
Protohymenopterdy 203 
ProtolabiSy 262 
Protolindeniciy Fig. 106 
ProtomeryXy 263 
Protoperlaria, 205, Figs. 104, 105 
Protozoa, 19, 23-25, 27, 33, 36, Figs. I3A,B 
ProtylopuSy 260, Fig. 138 
PseudocynodicttSy 231, 233, Fig. 118 



Pscudosuchians, as ancestors of birds, 186-188, 
Figs. 93, 94; as ancestors of dinosaurs, 145 

Psilophyton, 295 

Psocoptera, 204, 205 

Pteranodon, 167, 168, 171, 172, Fig. 87 

Ptcraspis, 76, 77, 81, Fig. 42 

Pteraspids, 76, 78, 88 

Pterichthys, 80 

Ptcridosperms, 296, Fig. 152 

Pterodactyls, 166, 167, Fig. 85 

Ptcrodactylus, 167, 171 

Pterosaurs, 125, 186; described, 164-168; habits 
of, 168-172 

Ptcrygota, 201, 202, 206 

Pterygotus, 31, 57, Fig. 29 

Raasch, R. I., 56 

Raccoons, 229 

Radial evolution, 305 

Radiates, 62-72 

Radiolarians, 19, 24, 30, 33, Fig. 6 

Reed, Bill, 17 

Reindeer, 250 

Relicts, 304 

Reptiles, 22; characteristics of skeleton, 119, 
120; classification, 124-126; extinction, 161- 
163; flying, 14, 162, 164-172; marine, 14, 
153-163; phalangeal formula, 120; range, 
301, Fig. 154; temporal fenestrae, 124-126 

Rhachitomi, in 

Rhamphorhynchus, 166-168, Fig. 86 

Rhinoceroses, evolution, 256-258 

Rhynia, Fig. 150 

Richthojcnia y 304 

Robergia, 52 

Rodents, 223 

Romcr, Alfred S., 150 

Rucdcmann, Rudolf, 59, 192, 302 

Ruminants, described, 248-252 

Saber-toothed tigers, 233-235, 269, Figs. 121, 


Salterella, 193 

Sauripterus, 98, 100, 102, Fig. 54 
Saurischia, 130, Fig. 71; described, 130-136 
Saurolophus, 139 

Sauropods, 130, 146, 161; described, 134-136 
Sayles, Robert W., 65 
Scelidosaurus, 140 
Schuchert, Charles, 147 
Scorpion flies, 204 
Scorpions, 60, 88, Fig. 30 

Scott, William B., 22*9, 250, 261 

Scyphozoa, 24 

Sea cucumbers, 20, 66, 72 

Seals, 229 

Sea urchins, 20, 40, 71 

Seeley, H. G., 164, 165, 169-171 

Selenodonts, 244, Fig. 129 

Sertularia, 42 

Sesamodon, 211 

Seymouria y 121, Figs. 646,0 

Sharks, 74, 78, 97, 100, Fig. 39 

Shimer, Hervey Woodburn, 312 

Shrews, 221, 223 

Sidneyia, 58, Fig. 10 

Sigillaria, 296, Fig. 151 

Simiidae, 276, 285, 286 

Simpson, George Gaylord, 215, 218, 222 

Sinanthropus petyngensis, 280, 287, Fig. I48E 

Sinopa grangeri, Fig. 117 

Skeleton, calcareous, formation of, 34-39, 4* 
in corals, 62; in echinoderms, 66, 94; in ver- 
tebrates, 94; siliceous, formation of: in ani- 
mals, 39; in plants, 294 

Skull, bones, 104; fenestrae, 124-126 

Skunks, 229 

Sloths, ground, 266; described, 268, 269; tree, 

Smilodon, 234, 235, Figs. i2iC, 122 

Snakes, 119, 125, 158, 162, 164 

Sollas, W. J., 42, 115 

Sphaerocoryphc pseudohemicranium, 10 

Sphenodon, 122, 125, 162 

Spiders, 55, 61 

Spirorbis, 302, 304 

Spirula, 190, 191 

Sponges, 19, 20, 23, 24, 26-28, 30, 32, 33, 36, 
37, Figs. 2, 7, 8, 130 

Squama ta, 187 

Squids, 189 

Starfish, 20, 40, 66-68 

Stegocephalia, 117, 121; described, 110-115 

Stegodon, 279, Fig. 141 

Stegosaurs, 130, 140, 146, 151 

Stegosaurus, 140, Fig. 74; sacral "brain," 149, 
162; small brain, 148 

Stenodictya, Fig. 102 

Stenomylus, 263, Fig. 134 

Stensio, E. A., 79, 81 

Stereospondyli, 112 

Sternberg, Kaspar Maria, Graf zu, 53 

Stetson, Henry Crosby, 13, 78, 80, 90 

Stigmaria, 296 



Stoermcr, Leif, 55 

Stone flies, 205 

Strongyloccntrotus droebrachiensis, 10 

Strophalosidy 303 

Strut hiomimuSy 134, 150, 186 

StylonurttSy 58, Fig. 27 

StyracosauruSy 142, Fig. 73 

Suioidta, 244 

Sus scrofa, 245 

Swim bladder, 99 

Symmetrodonts, 216-219, Fig. no 

SymphysopSy 48 

Synapsids, 125, 126, 212, Fig. 66B 

Synaptosauria, 125 

Tabulata, 64 

Taeniodonts, 223, 226 

TaeniolabiSy 213, Fig. 108 

Tait, David, 78 

Talbot, Mignon, 131 

Tapirs, 266, 305; described, 240-242, Figs. 124, 


Tapir us, 241, Fig. I25F 
Tarsioids, 276; described, 277-279, 286, 288 
Tarsius, 277 
Taxonomy, 9-11 
Teleoceras, 258, Fig. 135 
Teleosts, 74, 75, Figs. 39, 40, 41 
Temnospondyli, in 
TetoniuSy 277, 278, Fig. 146 
Tetrabranchiata, 189, 193, 194 
Tetracorals, 63, 64, Fig, 31; range, 301, Fig. 

Tetrapods, bones of limbs, 97; oldest, 99, 108- 


Thallatosuchia, 160 
Theriodonts, 211, 212, 219 
Thelodusy 77, 78, 80, 81 
Therapsida, 211, 212 
Theropoda, 139, 145; described, 130-134 
Thorpe, Malcolm R., 246, 248 
Thrinaxodon, 211, Fig. 107 
Thrips, 205 

Thysanoptera, 204, 205 
Tillodontia, 223, 225 
Tillyard, R. J., 32, 200, 202 
Titanichthysy 81, 83, 106 
TomarctuSy 231 
Torticones, 194, 198, Fig. 100 
Toxodontia, 223 
Trachodon, 138, 139 
Traquair, R. H., 78, 80 

Trails, 7, Fig. i 

TriccratopSy 18, 141, 142 

Trichoptera, 207 

Tricohodonts, described, 214-216, 218, Fig. 109 

TrigoniaSy 257 

Trilobites, 23, 32, 33; habits, 50-53; crawling, 

Figs. 23, 24; feeding, Fig. 25; molting, Fig. 

22; swimming, Fig. 24; relationships, 50; 

structure, 47, 48 
TrinuclcuSy 52 
Trioracodon, 214, Fig. 109 
Trituberculates, 216, 219 
Tritylodon longaeus, 213, Fig. 108 
Turriliticones, 198 

Turtles, 124, 125, 160, 162, 304, Fig. 84 
Tylopoda, 252 
TyrannosauruSy 133, 134, 141, 148, 149 

UintacrinuSy 70 

Uintatheres, 226 

Undinay 100 

Ungulata, described, 239-252 

Urodela, 108 

UrsuSy 231, 232 

VaranuSy 159 

Vermes, 20, 23, 24 

Vertebrae, embolomerous, in, Figs. 59A-C; 
lepospondylus, 113; phyllospondylus, 113, 
Fig* 571 double-ringed reptilian, 122, Figs. 
59E,F; rhachitimous, in, Fig. 590 

Vertebrates, 22, 73; characteristics, 75; emerge 
from water, 96; formation of skeleton, 94; 
origin: anaspid theory, 88-93; amphioxus 
theory, 84; annelid theory, 86; arthropod 
theories, 87; chordate theory, 84 

Volborthcllay 193 

VulpcSy 231 

Walcott, Charles D., 24, 25, 29-32, 34, 35, 58, 


Walruses, 229 
Wapiti, 250 
Wapticiy Fig. 10 
Wasps, 208 

Watson, D. M. S., 92, 93, 100, 105, no 
Weasels, 229 
Westoll, J. S., 104 
Wctmore, Alexander, 174, 180 
Whales, 227 

Wheeler, William Morton, 208 
Whittard, W. R, no